Final Report Summary - COSMOS (Cost-driven Adaptive Factory based on Modular Self-Contained Factory Units)
Executive Summary:
Wind turbines and aeronautic are sectors where Europe is leading the world class competition. To guarantee this competitiveness, their productivity has to be improved. The adoption of an automation strategy is a key factor to increase productivity. In the wind turbine manufacturing, the assembly is one of the core processes. Most of these assembly operations are manual which has the quality of being the most flexible way to do the work. The main objective for COSMOS is the design/development/implementation of a control system for factory management with a flexible, modular and evolvable automation approach which will permit to increase the assembly factory productivity by 20% without losing flexibility, focused on wind turbine assembly process although the solution will be suitable for other sectors. Cost models will be defined to assist in establishing the economically optimum factory’s configuration and automation level.
The achievement of the main objective will be obtained by fulfilling the following technical objectives:
• Create a factory organisation conception based on intelligent factory units for facilitating the self-adaptation to production changes under a flexible and modular automation configuration basis.
• Develop the distributed control system architecture according to such factory organisation.
• Develop the service layer infrastructure between control system and equipment involved in production.
Project Context and Objectives:
The main objectives for COSMOS are the design, development and implementation of a distributed control system for the management of a factory with a flexible, modular and evolvable automation approach capable of increasing the assembly factory productivity by 20% without compromising on flexibility. Although the project will focus on the wind turbine assembly process, the conception of the solution will be suitable for other sectors including aeronautics and railway industries.
One of the main drivers to increase the productivity is to introduce lean principles in these assembly processes integrating a control strategy that supports these principles. This approach will have an immediate impact, reducing reworks, waiting times (also for finished products), work in progress and the level of inventories.
Another project objective is the definition of cost models to assist in establishing the optimum configuration of the factory and its automation level from an economically advantageous perspective whilst taking into consideration evolvability, i.e. the capability of considering possible future scenarios.
The main objectives will be achieved by satisfying the following technical objectives:
• Create a factory organisation concept based on intelligent factory units facilitating self-adaptation to changes in product/production on a flexible and modular automation configuration basis.
• Design and develop a distributed control system architecture, which will take into account the conceived factory unit organisation and its modular configuration, and, additionally, will be capable of synchronising different factory units.
• Design and develop the service layer infrastructure between the control system and the equipment/devices involved in the production activities. This service layer will enable:
1. The interoperability with new devices in the modular factory units in a seamless, plug-and-use way, thus, facilitating the scalability of the solution.
2. The integration of task-oriented, automated sub-systems as it is envisaged to be the case for a variety of complex assembly, testing and inspection operations.
• Define cost models to optimise the factory unit configuration in terms of automation and modularisation. Such models will consider also the potential of adaptation capabilities with regards to new product demands.
The COSMOS system will have the following features:
1. Autonomous behaviour. The control system will support the autonomous behaviour of the factory units conceived.
2. Multilayer decentralised control. The control will work in three adjacent levels based on heterarchical and cooperative principles, i.e. factory, factory unit and process-device level.
3. Interoperable connectivity with factory units’ equipment/devices. Devices will be connected “plug-and-use”, thus facilitating the scalability of the system. Interoperability standards will be included for communication among systems and/or devices. Some examples of these “plug-and-use” functions are: identification, traceability and audit control.
4. Local, agent-based intelligence for process oriented self-adaptation. The factory unit control includes some task-oriented sub-systems as aggregation of devices with self-adaptation capacities (eliminating the need for human intervention). The local adaptive control system will facilitate the collaboration among equipment/devices to complete specific tasks. COSMOS will consider three main types of processes:
(1) Autonomous assembly operations with adaptive control to consider variable process requirements,
(2) Assembly testing with learning capacities for assembly validation criteria,
(3) Adaptive inspection for final checking.
COSMOS solutions will be implemented as a pilot installation for the assembly of wind turbines nacelles, the most critical turbine element with the highest potential for added value. Final assemblies can be up to twelve meters long and comprise thousands of parts with a high number of variants continuously evolving. Nowadays, the assembly process is carried out entirely in a manual way and on-site (the product is fixed in place and trained technicians carry out the required tasks). An automation solution requires the analysis of the best topology optimising the assembly flow. Figure B1.2 represents an initial idea of the assembly stages and flow depicting the envisaged results of the COSMOS project. This initial concept will need to evolve through a detailed analysis of the assembly process as part of the project.
The following sections describe the way that the COSMOS objectives are going to be addressed and implemented in order to provide a generic solution which is applicable to the particular scenario and, more generally, to a wide range of assembly processes.
Self-adaptivity – generic solution
As part of an assembly process a number of tasks that run on a set of collaborating hardware units, such as sensors and actuators, need to be performed in a self-adaptive way. The specific requirements analysis (see previous sections) shows the similarity and overlap between individual tasks of the assembly process and provides a base for creating a generic solution incorporating self-adaptivity.
The parameters/variables can be grouped together and treated in a general way, Fig. B1.9. The main modules, Sensing, Reasoning, Control and Acting, can be described in a general, conceptual way and their instantiation for the particular case is expected to be achieved through adaptation of the generic module. Self-adaptation is introduced in the form of a self-learning module that allows generalisation from specific models that describe the (physical or learned) behaviour of relevant aspects of the task.
It is envisaged that generic (conceptual) solutions will employ a high level of intelligence in order to perceive spatial and force tool-object relationships, adapt actions compensating for environmental changes and realise recovery from failure during the execution of the required tasks. The generic solution will provide mechanisms that allow optimising the task strategy in view of the environmental situation and the employed tools. The system will be able to cope with task uncertainties as well as changes in the environment and even physical changes of the tool itself. Dealing with uncertainty is a crucial key to efficiency when performing interaction between a tool and components that are possibly not exactly at a predefined location or whose dimensions deviate from the expected default.
Fig. B1.9: Generic self-adaptive sensing-reasoning&control-acting architecture
A number of different self-adaptive techniques will be studied. The main requirement for these techniques is to be able to adapt to manufacturing changes, preferably through learning, in order to carry out different tasks as part of an assembly process. Adaptation to changing tasks and conditions as perceived by sensors can be achieved through intelligent technologies – existing ones as well as novel ones to be created through the integration of interdisciplinary research as part of this project.
The following key adaptive methods will be investigated. Closed-loop hybrid (force and position) control based on feedback from multiple sensors (including vision, force and torque sensors) will be employed. A number of general classes of adaptive control can be considered; candidates include model-reference adaptive control, model predictive control, sliding mode adaptive control, dual control, self-organizing control, adaptive control using neural networks and fuzzy logic, non-linear control, and stochastic adaptive control.
Considering that self-adaptivy components are integrated in the full hierarchical control system, some difficulties are envisaged to carry out stability and performance analysis, going beyond the solutions provided by Lyaponov theorems. A number of approaches will be considered in this sense:
1. One approach is to model the system as a large scale system. When considering fuzzy-based approaches for modelling, these fuzzy models offer a systematic tool describing a large scale system formed by a number of local/small systems interconnected with each other. Lyaponov stability theory can then be applied for stability analysis. Part of the project will need to explore the feasibility of modelling the overall control system as a large scale fuzzy model (Wang, W.J. and Luoh, L. 2004).
2. Another approach is to employ fuzzy hierarchical modelling and control to deal with hierarchical systems (Wang, L.X. 1998) (Tsekouras, G. and Sarimveis, H. and Kavakli, E., Bafas, G. 2005). A nonlinear system formed by a connection of different modules in a hierarchical structure can be modelled by a hierarchical fuzzy model. Based on the hierarchical fuzzy model, fuzzy controllers can be designed under the consideration of system stability.
3. The stability of individual modules does not necessarily guarantee the stability of the overall system, especially, when considering cases where the modules are connected in a feedback loop. However, if it is possible to guarantee that the inner feedback control sub-systems are all stable, overall systems stability is implicitly shown. Anyway, the system's performance may need to be fine-tuned in this case.
4. Generally, quite a number of inroads have been made in recent years showing that Lyapunov-theorems-based stability can be shown also for complex, non-linear systems including neural networks and fuzzy-based systems. Please see references below (Lam H.K. 2005, Lam H.K. 2008, Lam H.K.2009).Adaptivity in sensing will be achieved through intelligent methods that will recognise and distinguish salient features in the acquired sensor data. Here a number of techniques will be considered to achieve this target. Model-based approaches can be used to predict ideal process/task signatures from measured position sensor signals and material knowledge; deviations from modelled behaviour will feed into the control strategy. The identification/optimisation of model parameters/variables can be achieved through a training procedure. Artificial Intelligence (AI) approaches involving learning strategies will also be considered. Neural Networks, Fuzzy-based methods and Genetic-Algorithm based methods will be investigated in this context. Learning will be achieved using appropriate training algorithms and features extracted from experimental data as well as input from experienced users. The generalisation capabilities of these AI-based methods will provide the required adaptivity which has the potential to be further modified through on-line learning. Further, the on-line and automated interpretation of sensor signals will enable real-time error detection and support for error recovery.
This project aims at automating these operations employing a generic framework of multi-agents. The solution will be applied to the specific scenarios, but it will be applicable to a wide range of assembly operations. The project will study agent-based methods in order to turn traditional, hard-wired operations into modern, adaptive ones capable of coping with the requirements of flexibility and parts uncertainty. Agents in the context of this proposal are sensing, actuating and reasoning/controlling devices with behaviours that can also be modelled as agents (see Fig. B1.9).
An agent-based approach has many advantages. Its modular structure allows reuse of agents for different tasks and applications. Because an agent-based approach is decentralised, short response times can be achieved providing real-time capability. Independence of agents leads to an increased robustness of the overall system in the face of failure locally. In a distributed implementation of multiple agents, self-adaptation can be enhanced through inter-agent communications relaying environment changes to other agents. Building on the proposed agent-based framework, agents can be created to provide self-adaptive solutions for assembly, testing and inspection. It is expected that a number of agents (such as, for example, force and torque sensor agents and actuation agents) can be reused across different operations.
Self-adaptivity – Specific application scenarios
The proposed agent-based concept will be applied to the three specific scenarios discussed above, Hub-bearing assembly, Yaw testing, Upper housing inspection. Agents will be formulated based on the general “sense-reasoning&control-acting model” and the specific requirements for each case (see Fig. B1.9 Table B1.1 and the Requirement analysis section). Self-adaptivity is achieved by providing each agent with an appropriate cost function and error metrics for the solution of the task; an optimisation step will enable the agent to complete a task successfully despite environment changes and uncertainty. Where needed, closed-loop hybrid (force and position) feedback control using multiple sensor inputs from vision, force and torque sensors will be employed. Adaptivity in vision will be achieved through intelligent methods recognising image segments that constitute specific, task-dependent features despite uncertainty in position and noise in camera images; input from CAD will be incorporate where possible. Integrating sensor signal input (such as force/torque sensors) and material/environment knowledge with model-based approaches, deviations from modelled behaviour will be predicted and used to feed into the adaptive control strategy. The identification/optimisation of model parameters can be achieved through a training procedure based on Artificial Intelligence (AI) approaches will be investigated in this context. Learning will be achieved using appropriate training algorithms and features extracted from experimental data as well as input from experienced users. The generalisation capabilities of these AI-based methods will provide the required adaptivity which has the potential to be further modified through on-line learning (self-adaptivity). The automated on-line interpretation of sensor signals will enable real-time error detection and error recovery.
PROCESS VARIABLES / PARAMETERS SELF-ADAPTIVE STRATEGIES
Assembly
(Hub-bearing Assembly) Force/time
Bolt elongation
Bolt material properties
Bolt position
Lack of parallelismOvalityCAD models - Adaptive control including Non-linear control, Neural network / Fuzzy logic based control, Model-reference and -predicive control
- Intelligent image recognition and analysis tools
- Optimisation techniques including Newton Raphson, Least square
- Learning methods (supervised, unsupervised)
Functional Testing
(Yaw testing) Yaw friction
Yaw vibration
Radial clearance
Hydraulic escapes - Emphasis on sensing, less on control/actuation
- Model-based methods for prediction of dynamic yaw behaviour
- Predict in field performance
- Fuzzy logic to model vague, human expert knowledge
Inspection
(Upper housing inspection) Vision
Surface laser profiling (3D)
Distances (ultrasonic sens) - Emphasis on sensing, less on control/actuation
- Sensor fussion
- Intelligent image recognition and analysis tools
- Intelligent adaptation to variability in product dimensions
Project Results:
Market (and stakeholder) analysis
This chapter contains a market analysis of the segments related to the exploitable results derived from COSMOS. Only select markets are analysed this way as some of the results have overlapping segments and some results are academic and as such usable in further research and not as a product.
This analysis explores the possible chances of products or services derived from COSMOS results on the open market and are used as a basis for further exploitation plans.
2.1 MES
MES is emerging as a superior product for flexible and integrated production. An MES combines all the properties required for modern production management in a comprehensive software system. This system is the basic and operational platform for all the enterprises evolving towards the digital factory of the future. MES solutions are increasingly used and recognized in the process manufacturing and discrete manufacturing industries (automotive, healthcare, aerospace and defence, FMCG etc.). This trend is accelerated by the ongoing economic globalization that requires maximized competitiveness.
The Global Manufacturing Execution System market will have a growth rate of 11.7 percent over the period 2011-2015. One of the key factors contributing to this market growth is the increasing demand from discrete manufacturing industries. Further, the Global Manufacturing Execution System market has also been witnessing emerging demand for MES solutions for mobile computing devices.
North America and Europe account for similar market share in the global MES market. Since the APAC region is becoming the manufacturing hub for many industries, this region possesses great potential for the MES market and most of the major vendors are focusing APAC now. Clearly, there are many vendors that offer MES software, but a big chance still exists out of question because of the potentiality and expandability of this market.
In the field of MES we differentiate between three topics: factory control, automatic manufacturing system and interoperability middleware, which gets its own chapter because of its importance.
2.2 Robotics and Automation
The situation concerning the production sector of wind turbines towards the end of the COSMOS project is assessed by Reis Robotics as robot manufacturer and system integrator as follows:
In the current period the investment rate in the wind turbine sector is rather low. Especially in Germany, this is mainly related to the existing over capacity of the amount of current being produced by the installed wind turbine plants. This over capacity is mainly resulting from a lack of power lines to convey the produced current from the production sites to the main areas of the power consumption. In many cases, investors for wind turbine installations postpone their activities until there will be a solution of the current transportation problem. This situation is especially valid for Germany and may be similar in other European countries. The situation all over Europe, however, seems to be heterogeneous.
Currently it can be stated from the Reis point of knowledge that the amount of realised wind turbine systems is rather reduced than expanded. Also the actual construction activities of new wind turbine plants are rather low, due to the described situation.
The COSMOS results are in the context of the project exemplarily demonstrated in the specific production step “bolt tensioning at the rotor hub of a wind turbine”. Principally, the self-adaptive production principles could also be applied to other production steps of the wind turbine production. Even other production scenarios in the component production, predominantly in the production of components in the renewable energy sector, could be addressed by the COSMOS results, as e.g. the battery production that becomes important in relation to this sector in future. A further innovative sector that can be addressed by the results could be the e-mobility area, especially in the production of lightweight parts for e-cars.
The sketched facts show that in any case there are several possibilities for a sustainable exploitation of the COSMOS principles of self-adaptiveness in practice in future times.
2.3 Interoperability Middleware
2.3.1 Communication interfaces of devices
Standard automation devices like PLCs offer different communication interfaces today. One group is field busses mainly used to connect I/O modules to the PLC but they are also used to connect the PLC to process management level. Another group is vendor specific TCP/IP based protocols. Both groups either communicate byte blobs without any type information or a list of variable values. OPC Data Access servers are available for all standard automation devices. They encapsulate the proprietary communication protocols in a standard interface on Windows PCs. OPC Data Access servers also limit the interface to a list of variables with values.
Special devices provide either serial connections or proprietary TCP/IP protocols. They often offer libraries for application programmers on Windows PCs.
MES protocols cannot be implemented on embedded devices since they are product specific and resource consuming. In addition most embedded device vendors do not allow to put any external components on the device.
2.3.2 Communication interfaces of MES systems
MES servers typically provide a vendor specific interface with a list of services. The communication is done through Enterprise Service Bus (ESB), web services, CORBA or vendor specific HTTP based protocols. Some vendors provide libraries for application programmers on Windows PCs.
Device protocols can normally not be implemented on a MES server since they communicate through the separation of process and IT networks.
2.3.3 Today’s integration strategy
The communication integration between devices and MES is typically done on line controller PCs or communication PCs in the production environment. On these PCs standard software of the MES is used to map supported protocols like OPC or Modbus TCP to the services and communication protocols used by the MES. If the device protocol is not supported, project specific software is developed that maps the device specific protocol to the MES services.
2.3.4 Status OPC UA integration on device level
OPC UA server functionality will be standard on PLCs. It is already supported by several vendors and much more have support announced. A major step will be support by the biggest vendors Siemens and Rockwell. Support for OPC UA client functionality was announced by several vendors and is important for vendors to allow device to device communication. The full functionality required by COSMOS is not supported yet.
Several machine builders have added OPC UA as standard interface to their products. Device vendors see OPC UA as the fallback communication channels to PLCs and other systems if the supported field bus is not used in the project. Prototypes exist but not many products.
The first step would be to integrate the necessary OPC UA functionality required by COSMOS OPC UA device service bus in a relevant number of devices. This requires updates of the device firmware. Standardized plug-and-use functionality can be added on a configuration level in most devices.
2.3.5 Status OPC UA integration on MES level
MES vendors like SAP and iTAC have integrated OPC UA into their products already. The level of support does not cover all aspects of the COSMOS OPC UA device service bus but there is a very high interest of MES vendors to standardize integration with the process layer. Therefore the concept of the COSMOS OPC UA device service bus is very interesting for MES vendors. Level of support will depend on level of support in devices. The first step is to integrate all necessary OPC UA features into MES systems.
The second step is the further standardization of the semantic of plug-and-use functionality. This will further simplify integration of process data with MES.
3 Positioning of COSMOS in standardization bodies
Standardization is one of the key success factors for COSMOS to reduce integration effort and costs for devices, subsystems and modular factory units. Therefore ASCOLAB is active in several different standardization working groups that focus on standardizing integration of different systems from the device level up to the enterprise level.
3.1 OPC Unified Architecture Working Group
The OPC Unified Architecture (UA) working group is a working group of the OPC Foundation. This is a permanent working group that defines the basic OPC UA functionality in a multipart specification with currently 13 parts. The results of this working group are also distributed through a liaison with IEC as international standard IEC 62541. Parts 1 – 10 are already available as IEC versions.
ASCOLAB is an active member of the working group since its start in October 2003. ASCOLAB is as editor responsible for two parts of the specification, Part 4 – Services and Part 9 – Alarms and Conditions.
OPC UA specification version 1.02
During the COSMOS project the OPC UA working group worked on a new release of the OPC UA specification with version number 1.02. Most of the updates released with this version in August 2012 were mainly maintenance enhancements and clarifications. Only a few new features were added to the existing parts. Two new parts were released in this version, Part 11 – Historical Access and Part 13 – Aggregates.
OPC UA Part 12 - Discovery
Another new specification, Part 12 – Discovery is relevant for COSMOS and defines enhanced discovery functionality necessary for the OPC UA device service bus in COSMOS. This specification part is not released yet since it requires prototype implementation and detailed evaluation. A first release candidate for final review and voting will be available in December 2013.
The two relevant features for COSMOS OPC UA device service bus in OPC UA Part 12 are the ad-hoc service discovery and the Global Directory Service (GDS).
Ad-hoc service discovery allows OPC UA servers to announce their service functionality in a local subnet using mDNS protocol. OPC UA clients can find OPC UA servers and their service capabilities by accessing a local mDNS protocol cache in the Local Discovery Server (LDS). ASCOLAB implemented two prototypes of the mDNS protocol integration into the LDS. This multicast extension to the LDS is used by servers to announce their functionality in the network and by clients to access the local mDNS cache. The second prototype will be the base for the new LDS-ME (Local Discovery Server with Multicast Extension) provided by the OPC Foundation as standard discovery component.
GDS is a standardized OPC UA interface that encapsulates a system for plant wide OPC UA application configuration and discovery. The discovery functionality extends the subnet discovery through mDNS protocol and provides a global service registry that works plant or corporate wide across different subnets. The second part of the interface allows a central configuration of application level security by managing and distributing security certificates and application trust relations. The interaction between the GDS and OPC UA applications is defined in Part 12. The internal implementation of a GDS can be a wrapper around an existing system like a Windows Active Directory or a stand-alone application for central OPC UA device service bus configuration. ASCOLAB developed a GDS prototype for evaluation in the OPC UA working group.
3.2 OPC Foundation and PLCopen joined working groups
PLCopen is the organization that maintains the international standard IEC 61131-3. This standard defines PLC programming languages and a software model for PLCs. The joined working groups between OPC Foundation and PLCopen define essential functionality necessary to fully integrate off the shelf PLCs into the OPC UA device service bus.
OPC UA Information Model for IEC 61131-3 (PLCopen)
This working group defines the details for integration of an OPC UA server into PLCs. ASCOLAB is an active member of the working group as editor of the specification and joined all working group meetings. The first version of the OPC UA Information Model for IEC 61131-3 defines a mapping of the IEC 61131-3 software model to OPC UA objects and covers mainly the data access use case. The first version is released since March 2010
The mapping of additional OPC UA features like Methods and Events is currently under definition in version 2.0. These features are required for integrating PLCs like the one from PHOENIX CONTACT with all features into the OPC UA device service bus for COSMOS. ASCOLAB presented use cases to the working group.
OPC UA communication function blocks for IEC 61131-3
This working group defines details for integration of an OPC UA client into PLCs. ASCOLAB is an active member of the working group as OPC UA expert and joined all working group meetings.
The working group defines function blocks that allows a PLC programmer to act as OPC UA client from within the PLC program. All OPC UA client functionality necessary for COSMOS is covered by the first specification version. The functionality is an essential part for the integration of PLCs into the OPC UA device service bus.
The first results of the working group with a working prototype between a Beckhoff PLC acting as OPC UA client and the MES system from iTAC acting as OPC UA server were shown at the Hannover Fair in spring 2013. The two PLC vendors Beckhoff and B&R Automation finished a complete prototype implementation of the defined function blocks in September 2013. Based on the prototyping results a release candidate specification is planned for December 2013.
3.3 MES Connectivity
This OPC Foundation working group was started in December 2011 together with PLCopen, MESA (Manufacturing Enterprise Solutions Association), VDMA (German Machine Builder Association), ZVEI (German Electro Technical Industry Association) to standardize integration of devices and MES based on OPC UA. The requirements for the OPC UA device service bus in COSMOS are completely covered by the use cases defined by this working group.
ASCOLAB is one of the initiators of the working group and active as editor of the working group.
3.4 OPC DI version 2.0 and FDI (Field Device Integration)
Field Device Integration (FDI)
The FDI working group defines an international standard for configuration and maintenance of field devices. The working group consists of the standardization groups PNO (Profibus User Organisation), HART Foundation, OPC Foundation, FDT Group (Field Device Tool), Fieldbus Foundation and EDDL (Electronic Device Description Language). FDI will integrate and harmonize all of these standards and will integrate them into one common standard.
ASCOLAB is member of the review team and the handheld working group for the new FDI standard.
OPC UA for Devices (DI)
OPC DI is the base OPC UA information model for FDI. Generic features for device configuration defined by the FDI working group were integrated into OPC DI version 2.0. This new specification version was released in July 2013.
ASCOLAB provides the chairman for this working group.
3.5 OPC UA Information Model for ISA95
The goal of this OPC Foundation working group is the definition of a mapping of the established standard ISA 95 to OPC UA. The working group was started in January 2012.
HOLOS and ASCOLAB are represented in the ISA95/OPC-UA Working Group. The goal is to develop the standard with the contribution of Cosmos requirements and use cases. This will enable the use of the standard in COSMOS. Using this standard in COSMOS enables the plug and use functionalities and integrates in COSMOS a standardized communication interface. This standard is already used worldwide through B2MML.
The specification was distributed as release candidate for voting to the OPC Foundation Technical Advisory Council (TAC) in October 2013. ASCOLAB as member of the TAC and a majority of members voted for the release of the specification. The distribution of the final release document is expected for November 2013.
3.6 AutomationML, IEC SC 65 E, Project no. 62714 Ed 1.0
The goal of this working group is to standardize engineering data exchange for use in industrial automation engineering.
EDAG is a Steering Committee member of the AutomationML group and active in the working group. EDAG has evaluated AutomationML for use in the COSMOS project.
4 Business model and COSMOS exploitation strategy
In this chapter business models will be discussed in greater detail. This includes existing business models of some partners involved in COSMOS, which are showcased here as an example. Additionally a more general approach of possible business models for exploiting the results of COSMOS is illustrated.
4.1 Existing business models of partners
In this chapter the business models of two COSMOS partners will be showcased. To describe these business models the CANVAS schema is used. For the sake of simplicity the usual CANVAS panel is transformed into an easily readable table.
4 1 CANVAS explanation
4.1.1 EDAG
Key partner - COPADATA
- FFT
- Employees
Key activities - Consulting for IT systems
- Problem support for IT systems
- Development of concepts for IT systems
- Development of IT systems
Key resources - Developers, project managers, consultants
- Development environments
Value proposition/services - Process improvements
- Automating through IT
- Analysis and reporting
- Process development
- Software development
Relationships - Technology partner: Support for partner applications, development using partner solutions
- Affiliated company: Temporary distribution of manpower, joint processing of inquiries
- Key user: improvement ideas, new inquiries
Channels - Inquiries from OEMs: distributed from company and affiliates
- Key-accountant: advertising services to OEMs
- Follow-up orders from regular customers
- Trade shows (both internally and externally organized)
Clients, customer segments - OEMs: consulting, support and development MES and ERP systems
- Segments: automotive, aerospace, rail, general assembly, medical and component suppliers
Costs, investments - Team: costs for team members (personnel, travel, facilities)
- Time: time employees and managers spend within workshops/projects.
- Development tools: license costs, hardware costs, administration costs.
- External consultants: subcontracting for specialized expertises
Revenue improvements, cost savings wins - Pricing: dependant on time invested (personnel and travel expenses included)
- Value: process improvements reduce running costs by reducing needed resources (time, personnel, equipment), quality is improved and
4.1.2 IBERMATICA
In this subchapter the business model of IBERMATICA is described using the CANVAS-schema in tabular form.
Key partner - PARTNERS: Microsoft, IK4 – TEKNIKER
- Quality -> CMMI
- Providers: PENTA, SIEMENS, ROCKWELL, MATRIKON
- ERP providers: SAP, AXAPTA, NAVISION, RPS, ...
- Employees! Key driver/initiators for improvement
Key activities - Process improvement consultancy/service
- Problem solving support
- Process monitoring, decision system support, taceabilty
Key resources - Project managers, MES consultants, quality consultants
- MES modelling software, data mining
Value proposition/services - Real estate plant monitoring
- Presence control, work oder tracking, real downtime control
- Management, documentation, production support
- Maintenance management
- KPI analysis and reports
- Support for continuous improvement process
Relationships - Process council: Meetings of top management with process owners and managers in order to improve processes
- IT & Quality: Support from IT for process automation. Support quality to achieve ISO requirements.
- Technology partner: Customer support in the application of new technologies to manage production
Channels - Cross selling: ERP Solutions, CAM, quality, maintenance
- Process improvement projects: Strategic projects to redesign/heavily improve processes
- Communication via multiple channels: Intranet, workshops, product presentations
Clients, customer segments - Upper management: Provide tools to improve process performance.
- Middle management: Driving operational effiency (more with less).
- Employees: Making their daily job easier by providing a system to capture process data.
- Segments: Automotive, aerospace, food, plastic injection, manual assembly.
Costs, investments - MES Team: Costs for MES team members (personnel, travel, facilities)
- Time: Time employees and managers spend within workshops/projects.
- OPC, .NET tools: License costs, hardware costs, administration costs.
- MES consultants: Bringing in external consultants to support with specific expertise.
Revenue improvements, cost savings wins - Pricing: Free of charge for internal improvements. Consultancy fee for external improvements.
- Value: Increased revenue from improved processes. Reduced lead time. Reduced costs by improving quality & speed.
4.2 Possible business models
This chapter explains different general business models relevant to the COSMOS partners. While not all models are applicable to every result, these present possible approaches to facilitate the exploitation of the results.
4.2.1 Service-oriented
A service-oriented business model concentrates on delivering a certain service, for example consulting, usually for a hourly or daily fee. In this model, the knowledge and experience of the employees is the greatest asset and revenue is generated by using this asset to improve the processes of the client.
Key partner - Consultants: provide knowledge and experience
Key activities - Consulting
- Support
Key resources - Employees
- Knowledge
- Experience
Value proposition/services - Consulting improves processes
- Support to find errors and minimize outages
Relationships - Technology partners: Support for certain technologies used
Channels - Trade fairs
- Technology partners
Clients, customer segments - Businesses using the tools/processes the employees are experienced with
Costs, investments - Employees
- Equipment and tools for the employees
Revenue improvements, cost savings wins - Time based fees for consultation and support.
4.2.2 Solution provider
A Solution provider is contracted to find a solution to a certain problem. The solution itself can be a process, a tool or a concept to counteract the problem. Revenue is generated by selling the solution, usually at a price determined at the beginning.
Key partner - Consultants: provide knowledge and experience
Key activities - Development of tools/processes/concepts
Key resources - Employees
- Knowledge
- Experience
Value proposition/services - The problem is solved for the client or an existing process was optimized and is now faster/cheaper/improved
Relationships - Technology partners: Support for certain technologies used
Channels - Trade fairs
- Technology partners
Clients, customer segments - Businesses wanting to solve a certain problem hindering their productivity
Costs, investments - Employees
- Equipment and tools for the employees
Revenue improvements, cost savings wins - Contract fee for the solution as a whole
4.2.3 Component provider
A component provider supplies a finished product to customers in need of a certain tool or resource. Supporting and marketing this product is usually also part of this business model. Revenue is generated by selling units of the product or licenses allowing its use.
Key partner - Marketing partners
- Distributors
- Possibly subcontractors
Key activities - Marketing and supporting the product
- Further development of the product
Key resources - Employees
- Experience
- Resources needed to create the product
Value proposition/services - The product itself as a tool or component needed to create another product or a service
Relationships - Technology partners: Support for certain technologies used
- Marketing: Generating a market for the product
- Customers: Feedback for product improvements
Channels - Trade fairs
- Technology partners
- Trade with contractors and/or subcontractors
Clients, customer segments - Customers in need of a certain tool or resource fulfilled by the product
Costs, investments - Employees
- Equipment and tools for the employees
- Resources to create the product
- Further development of the product
Revenue improvements, cost savings wins - The product as a tool or resource to generate revenue or fulfil a certain need
4.3 COSMOS general exploitation strategy
This chapter contains information about the general strategy used as a basis for the detailed exploitation strategies. The possible business models serve as a context in which the results can be exploited.
The final results of the COSMOS project depicted in the figure below are referenced by the table following it. This table gives an overview over the exploitation plans detailed later in the document and states the business model(s) to be associated with every exploitation strategy.
4 2 Overview COSMOS final results
N° Exploitation strategy Final result Relevant business model Owner
1 Automation ML Connector for MES (AML2MES Adapter)
Automation ML Connector Component provider and service-oriented EDAG
2 Registration and CAD2Robot Trajectory Planning
Robot Based Inspection Solution provider KCL
3 Self-adaptive Bolt Tightening
Bearing Assembly Solution provider KCL
4 Self-adaptive Testing
Computer Based Texting Solution provider TEKNIKER
5 Self-adaptive Assembly System
Bearing Assembly Solution provider REIS
6 Factory Control/SCFU Control
Factory Control Solution provider IBERMATICA
7 Interoperability Middleware
Interoperability Middleware Component provider and service-oriented ASCOLAB
Potential Impact:
In this section the exploitable results of the Cosmos project will be briefly described, also highlighting the main progress beyond the State of the Art. The main “owner” for each Cosmos component will be identified and the owner/s will be the main responsible for the exploitation of the specific part of the Cosmos project.
The following table reports on exploitable results of the project which the partners intend to exploit.
N° Result Has the result commercial/ social significance? Can the result be exploited as a stand alone product, process, service, etc? Should the result be grouped with one or more others? Should the result be split into one or more others? Reporter
1.0 Factory Control IBERMATICA
1.1 Factory control: Lean production support Yes No Yes No
1.2 SCFU workflow definition Yes No Yes No
1.3 SCFU control. MES components Yes No Yes No
1.3.1 OPC/UA Ibermatica Yes No Yes No
1.3.2 OPC/UA EDAG Yes No Yes No
1.3.3 Automation ML adapter EDAG Yes No Yes No
1.4 SCFU control: HMI content editor Yes No Yes No
2.0 Flexible automatic manufacturing system REIS ROBOTICS
2.1 Self-adaptive control algorithms Yes Yes No Yes KCL
2.2 (Self-adaptive) geometrical verification system Yes Yes Yes No stt, KCL
2.3 Self-adaptive assembly process Yes Yes Yes No ZEMA, REIS ROBOTICS, KCL
2.4 Self-adaptive inspection process Yes Yes No No
2.5 Economic model and evaluation tool Yes Yes Yes No RWTH WZL
3.0 Interoperability Middleware
3.1 Automation ML connector for MES Yes Yes No Yes EDAG
3.2 OPC/UA connector for Phoenix PLC Yes Yes No Yes ASCOLAB
3.3 OPC/UA connector for MES Yes Yes No Yes ASCOLAB/ IBERMATICA/ EDAG
3.4 B2MML-based connector for MES/ERP Yes Yes No Yes IBERMATICA
3.5 Knowledge on plug and use functions/standardisation Yes Yes No Yes ASCOLAB/ HOLOS
3.6 OPC/UA connector for OPC data access Yes Yes No Yes ASCOLAB
Table 3 1 Exploitation results
3.1 Market Analysis
This chapter contains a market analysis on three major topics COSMOS is dealing with.
The analysis contains a description of the innovations the products brings, an outline of potential customers and what benefits the customers may expect. In addition to that, dates for the estimated timeline to market the product are given and costs for the exploitation are summarised. Furthermore the potential product price range and market volume are estimated. Then, a list of competitors is compiled and an explanation on how the new product will rank against their products and how fast they are expected to react. Finally all involved partners are listed and the modalities of protecting the product are named.
This chapter contains a market analysis on three major topics that COSMOS is dealing with. All the three parts are based on Self-Contained Factory Unit (SCFU) and are important elements of a Manufacturing Execution System (MES).MES is emerging as a superior product for flexible and integrated production. An MES combines all the properties required for modern production management in a comprehensive software system. This system is the basic and operational platform for all the enterprises evolving towards the digital factory of the future. MES solutions are increasingly used and recognized in the process manufacturing and discrete manufacturing industries (automotive, healthcare, aerospace and defense, FMCG etc.). This trend is accelerated by the ongoing economic globalization that requires maximized competitiveness.
The Global Manufacturing Execution System market will have a growth rate of 11.7 percent over the period 2011-2015. One of the key factors contributing to this market growth is the increasing demand from discrete manufacturing industries. Further, the Global Manufacturing Execution System market has also been witnessing emerging demand for MES solutions for mobile computing devices.
North America and Europe account for similar market share in the global MES market. Since the APAC region is becoming the manufacturing hub for many industries, this region possesses great potential for the MES market and most of the major vendors are focusing APAC now. Clearly, there are many vendors that offer MES software, but a big chance still exists out of question because of the potentiality and expandability of this market.
We differentiate the MES here in Factory control, Flexible automatic manufacturing system and interoperability middleware three topics.
Factory control
Describe the innovation content of result 1 Distributed Control Architecture in Assembly Processes, Factory and SCFU control = SW packages
Who will be the customer? Industry
What benefit will it bring to the customers? Modular FActroy Control System with new Standard communication links (OPC UA) and For- Backward data exchange with engineering systems.
Modular Manufacturing Execution System with Interoperability with ERP system.
Ease of handling by non expert users. Components for HMI creation and workflow definition
When is the expected date of achievement in the project (Mth/yr)? Aug 2013
When is the time to market (Mth/yr)? Jun 2014
What are the costs to be incurred after the project and before exploitation? 200 K€ (EDAG)
275 K€ (IBER)
What is the approximate price range of this result / price of licences? MES solution 25 k€
OPC UA 3k€
AutomationML 35 k€
What is the market size in M€ for this result and relevant trend? 25 M€/a (IBER) + 10 M€/a (EDAG)
How this result will rank against competing products in terms of price / performance? Us achieving this goal by developing a tool that allows end users to easily configure their own solution.
OPC UA – same price/performance
AutomationML no competing product
Who are the competitors for this result? (Fraunhofer IOSB), Siemens (EDAG)
SYSTEPLAN, IDS, MAPEX (IBER)
How fast and in what ways will the competition respond to this result? 3 to 5 Years
Who are the partners involved in the result? OPC UA: Ascolab,
Workflow definition : TEKNIKER
AutomationML (none)
Who are the industrial partners interested in the result (partners, sponsors, etc…)? Ibermatica, EDAG, Phoenix, Gamesa
Have you protected or will you protect this result? How? When? By the end of the project.
Table 3 2 Factory Control exploitation results
3.1.2 Flexible automatic manufacturing system
Describe the innovation content of result1 Result is a piece of machinery / a process / a method; within the context of a wind turbine production line; flexible production method (self-adaptive); new production method; new technology; strength: improved flexible production process
Who will be the customer? Customers from the wind turbine sector and/or also from other suitable industries.
What benefit will it bring to the customers? More flexible production process.
When is the expected date of achievement in the project (Mth/yr)? 05/2013 (approx. 3 months before project end).
When is the time to market (Mth/yr)? Project end + 2 years: 11/2015.
What are the costs to be incurred after the project and before exploitation? 6-12 person months (only rough guess).
What is the approximate price range of this result / price of licences? 100,000 Euros.
What is the market size in M€ for this result and relevant trend? 5 systems per year (5 x 100,000 Euros).
How this result will rank against competing products in terms of price / performance? Price: higher (20%); performance: better (30%).
Who are the competitors for this result? Other system integrators in the wind turbine production line sector.
How fast and in what ways will the competition respond to this result? 2–3 years after launch of COSMOS result.
Who are the partners involved in the result? Reis, STT, Phoenix, WZL, KCL, ZeMA
Who are the industrial partners interested in the result (partners, sponsors, etc…)? Reis, STT, Phoenix
Have you protected or will you protect this result? How? When? If possible and necessary: by patents. Currently open, whether the COSMOS results can/will also be patented.
Table 3 3 Flexible automatic manufacturing system exploitation results
3.1.3 Interoperability Middleware
Describe the innovation content of result1 Knowledge in operation and development of Interoperability Issues, integrating OPC/UA
Who will be the customer? Industry, Self.
What benefit will it bring to the customers? Pluggability infrastructure for different component types (software and hardware)
When is the expected date of achievement in the project (Mth/yr)? 1st Year
When is the time to market (Mth/yr)? Two years.
What are the costs to be incurred after the project and before exploitation? ~50 K€
What is the approximate price range of this result / price of licences? ~ 100 K€
What is the market size in M€ for this result and relevant trend? 3 M€
How this result will rank against competing products in terms of price / performance? Better, cheaper and faster regarding implementation costs.
Who are the competitors for this result? All companies implementing DPWS and OPC/UA.
How fast and in what ways will the competition respond to this result? 1 to 2 Years
Who are the partners involved in the result? Ibermatica, Tekniker, Ascolab, Holos
Who are the industrial partners interested in the result (partners, sponsors, etc…)? Ascolab, Reis, Holos, Edag, Phoenix
Have you protected or will you protect this result? How? When? By the end of the project.
Table 3 4 Interoperability middleware exploitation results
Various dissemination tools that have been developed by the project so far and outlines their purpose as means of sharing information about the project and generating interest in the work being undertaken by the COSMOS project.
3.1 COSMOS website
Target audience: Public, interested 3rd parties
A website has been created for the project and can be accessed using the following url: http://www.cosmosproject.eu/
As the project’s main dissemination channel, the website serves as a source of information about the project for the public and all external parties who may have an interest in the project. The maintenance and updating of the COSMOS website is done by IBERMATICA, project coordinator. However, the input for the content is generated by all of the project partners.
The website provides the following information:
• General static information about the project, including the project concept, partners, volume and contact information.
• News, events and press releases
• Publications; public deliverables and other publications as and when they are available
3.2 Dissemination material
3.2.1 Project Presentation & Leaflet
A Leaflet was prepared and publish with a brief project presentation in English, containing key information about the project (participants, approach, expected achievements etc.). In addition, this presentation was attached also on the project’s website and within other dissemination activities.
Figure 1 - Cosmos Brouchure A
Figure 2 - Cosmos Brouchure B
3.2.2 Roll-up posters
For a more effective means of visual communication and dissemination several roll-ups were designed and printed to be exposed in conferences and industrial fares and set up alongside the project exhibition booth.
3.2.3 Videos
Several videos were created during the project in order to expose advances.
3.2.3.1 Self adpative tensioning process at the rotor hub
Within the Cosmos Project a self-adaptive tensioning process and tool are developed which allow an exact preload application to the bolts of a wind turbine pitch bearing. In the first video the tensioning process is shown at the demonstrator which consists of six bolts and is used for testing the process and the tool. Moreover, the video presents the human interface to control the process and monitor different parameters. In the first step of the process the tensioning tool stops over the bolt. In the next step the tool is moved linear onto the bolt till it gets into contact with the surface of the bearing ring. If the contact is detected the change bushing is threaded onto the bolt and pressure is built up. In this condition the bolt is stretched and the nut is tightened. At the End of the process the pressure is released and the tensioning cylinder removed.
http://www.youtube.com/watch?v=LdGqWRqkcnU&feature=youtu.be
Figure 3 - Self adaptive tensioning process at the rotor hub
3.2.3.2 Self adaptive tensioning process at the demonstrator
The results of the initial tests of the demonstrator a used for the validation at real hub of a 2MW wind turbine. This video shows the first tests of the tensioning process at the real rotor hub. At the beginning the position of the bearing is detected. Therefore, the robot measures different points of the rotor hub with a laser sensor. The tensioning points of the bolts are calculated within the measured rotor hub coordinate system with the product parameters. So the tool is guided by the robot to the different tensioning points where the bolts are tensioned one by one.
http://www.youtube.com/watch?v=4WZGgnrr2PE&feature=youtu.be
Figure 4 - Self adaptive tensioning process at the demonstrator
3.2.3.3 COSMOS Geometrical Inspection
Self adaptive testing and inspection. The object inspection is performed with 3D laser scanner fixed on industrial robot. The surfaces are scanning according to CTQs (Critical To Quality) definition. The robot trajectory planning are originates from object CAD drawing and CTQs.
http://www.youtube.com/watch?v=IivHjThjm2g&feature=youtu.be
Figure 5 - COSMOS Geometrical Inspection
List of Websites:
www.cosmosprojec.eu
Wind turbines and aeronautic are sectors where Europe is leading the world class competition. To guarantee this competitiveness, their productivity has to be improved. The adoption of an automation strategy is a key factor to increase productivity. In the wind turbine manufacturing, the assembly is one of the core processes. Most of these assembly operations are manual which has the quality of being the most flexible way to do the work. The main objective for COSMOS is the design/development/implementation of a control system for factory management with a flexible, modular and evolvable automation approach which will permit to increase the assembly factory productivity by 20% without losing flexibility, focused on wind turbine assembly process although the solution will be suitable for other sectors. Cost models will be defined to assist in establishing the economically optimum factory’s configuration and automation level.
The achievement of the main objective will be obtained by fulfilling the following technical objectives:
• Create a factory organisation conception based on intelligent factory units for facilitating the self-adaptation to production changes under a flexible and modular automation configuration basis.
• Develop the distributed control system architecture according to such factory organisation.
• Develop the service layer infrastructure between control system and equipment involved in production.
Project Context and Objectives:
The main objectives for COSMOS are the design, development and implementation of a distributed control system for the management of a factory with a flexible, modular and evolvable automation approach capable of increasing the assembly factory productivity by 20% without compromising on flexibility. Although the project will focus on the wind turbine assembly process, the conception of the solution will be suitable for other sectors including aeronautics and railway industries.
One of the main drivers to increase the productivity is to introduce lean principles in these assembly processes integrating a control strategy that supports these principles. This approach will have an immediate impact, reducing reworks, waiting times (also for finished products), work in progress and the level of inventories.
Another project objective is the definition of cost models to assist in establishing the optimum configuration of the factory and its automation level from an economically advantageous perspective whilst taking into consideration evolvability, i.e. the capability of considering possible future scenarios.
The main objectives will be achieved by satisfying the following technical objectives:
• Create a factory organisation concept based on intelligent factory units facilitating self-adaptation to changes in product/production on a flexible and modular automation configuration basis.
• Design and develop a distributed control system architecture, which will take into account the conceived factory unit organisation and its modular configuration, and, additionally, will be capable of synchronising different factory units.
• Design and develop the service layer infrastructure between the control system and the equipment/devices involved in the production activities. This service layer will enable:
1. The interoperability with new devices in the modular factory units in a seamless, plug-and-use way, thus, facilitating the scalability of the solution.
2. The integration of task-oriented, automated sub-systems as it is envisaged to be the case for a variety of complex assembly, testing and inspection operations.
• Define cost models to optimise the factory unit configuration in terms of automation and modularisation. Such models will consider also the potential of adaptation capabilities with regards to new product demands.
The COSMOS system will have the following features:
1. Autonomous behaviour. The control system will support the autonomous behaviour of the factory units conceived.
2. Multilayer decentralised control. The control will work in three adjacent levels based on heterarchical and cooperative principles, i.e. factory, factory unit and process-device level.
3. Interoperable connectivity with factory units’ equipment/devices. Devices will be connected “plug-and-use”, thus facilitating the scalability of the system. Interoperability standards will be included for communication among systems and/or devices. Some examples of these “plug-and-use” functions are: identification, traceability and audit control.
4. Local, agent-based intelligence for process oriented self-adaptation. The factory unit control includes some task-oriented sub-systems as aggregation of devices with self-adaptation capacities (eliminating the need for human intervention). The local adaptive control system will facilitate the collaboration among equipment/devices to complete specific tasks. COSMOS will consider three main types of processes:
(1) Autonomous assembly operations with adaptive control to consider variable process requirements,
(2) Assembly testing with learning capacities for assembly validation criteria,
(3) Adaptive inspection for final checking.
COSMOS solutions will be implemented as a pilot installation for the assembly of wind turbines nacelles, the most critical turbine element with the highest potential for added value. Final assemblies can be up to twelve meters long and comprise thousands of parts with a high number of variants continuously evolving. Nowadays, the assembly process is carried out entirely in a manual way and on-site (the product is fixed in place and trained technicians carry out the required tasks). An automation solution requires the analysis of the best topology optimising the assembly flow. Figure B1.2 represents an initial idea of the assembly stages and flow depicting the envisaged results of the COSMOS project. This initial concept will need to evolve through a detailed analysis of the assembly process as part of the project.
The following sections describe the way that the COSMOS objectives are going to be addressed and implemented in order to provide a generic solution which is applicable to the particular scenario and, more generally, to a wide range of assembly processes.
Self-adaptivity – generic solution
As part of an assembly process a number of tasks that run on a set of collaborating hardware units, such as sensors and actuators, need to be performed in a self-adaptive way. The specific requirements analysis (see previous sections) shows the similarity and overlap between individual tasks of the assembly process and provides a base for creating a generic solution incorporating self-adaptivity.
The parameters/variables can be grouped together and treated in a general way, Fig. B1.9. The main modules, Sensing, Reasoning, Control and Acting, can be described in a general, conceptual way and their instantiation for the particular case is expected to be achieved through adaptation of the generic module. Self-adaptation is introduced in the form of a self-learning module that allows generalisation from specific models that describe the (physical or learned) behaviour of relevant aspects of the task.
It is envisaged that generic (conceptual) solutions will employ a high level of intelligence in order to perceive spatial and force tool-object relationships, adapt actions compensating for environmental changes and realise recovery from failure during the execution of the required tasks. The generic solution will provide mechanisms that allow optimising the task strategy in view of the environmental situation and the employed tools. The system will be able to cope with task uncertainties as well as changes in the environment and even physical changes of the tool itself. Dealing with uncertainty is a crucial key to efficiency when performing interaction between a tool and components that are possibly not exactly at a predefined location or whose dimensions deviate from the expected default.
Fig. B1.9: Generic self-adaptive sensing-reasoning&control-acting architecture
A number of different self-adaptive techniques will be studied. The main requirement for these techniques is to be able to adapt to manufacturing changes, preferably through learning, in order to carry out different tasks as part of an assembly process. Adaptation to changing tasks and conditions as perceived by sensors can be achieved through intelligent technologies – existing ones as well as novel ones to be created through the integration of interdisciplinary research as part of this project.
The following key adaptive methods will be investigated. Closed-loop hybrid (force and position) control based on feedback from multiple sensors (including vision, force and torque sensors) will be employed. A number of general classes of adaptive control can be considered; candidates include model-reference adaptive control, model predictive control, sliding mode adaptive control, dual control, self-organizing control, adaptive control using neural networks and fuzzy logic, non-linear control, and stochastic adaptive control.
Considering that self-adaptivy components are integrated in the full hierarchical control system, some difficulties are envisaged to carry out stability and performance analysis, going beyond the solutions provided by Lyaponov theorems. A number of approaches will be considered in this sense:
1. One approach is to model the system as a large scale system. When considering fuzzy-based approaches for modelling, these fuzzy models offer a systematic tool describing a large scale system formed by a number of local/small systems interconnected with each other. Lyaponov stability theory can then be applied for stability analysis. Part of the project will need to explore the feasibility of modelling the overall control system as a large scale fuzzy model (Wang, W.J. and Luoh, L. 2004).
2. Another approach is to employ fuzzy hierarchical modelling and control to deal with hierarchical systems (Wang, L.X. 1998) (Tsekouras, G. and Sarimveis, H. and Kavakli, E., Bafas, G. 2005). A nonlinear system formed by a connection of different modules in a hierarchical structure can be modelled by a hierarchical fuzzy model. Based on the hierarchical fuzzy model, fuzzy controllers can be designed under the consideration of system stability.
3. The stability of individual modules does not necessarily guarantee the stability of the overall system, especially, when considering cases where the modules are connected in a feedback loop. However, if it is possible to guarantee that the inner feedback control sub-systems are all stable, overall systems stability is implicitly shown. Anyway, the system's performance may need to be fine-tuned in this case.
4. Generally, quite a number of inroads have been made in recent years showing that Lyapunov-theorems-based stability can be shown also for complex, non-linear systems including neural networks and fuzzy-based systems. Please see references below (Lam H.K. 2005, Lam H.K. 2008, Lam H.K.2009).Adaptivity in sensing will be achieved through intelligent methods that will recognise and distinguish salient features in the acquired sensor data. Here a number of techniques will be considered to achieve this target. Model-based approaches can be used to predict ideal process/task signatures from measured position sensor signals and material knowledge; deviations from modelled behaviour will feed into the control strategy. The identification/optimisation of model parameters/variables can be achieved through a training procedure. Artificial Intelligence (AI) approaches involving learning strategies will also be considered. Neural Networks, Fuzzy-based methods and Genetic-Algorithm based methods will be investigated in this context. Learning will be achieved using appropriate training algorithms and features extracted from experimental data as well as input from experienced users. The generalisation capabilities of these AI-based methods will provide the required adaptivity which has the potential to be further modified through on-line learning. Further, the on-line and automated interpretation of sensor signals will enable real-time error detection and support for error recovery.
This project aims at automating these operations employing a generic framework of multi-agents. The solution will be applied to the specific scenarios, but it will be applicable to a wide range of assembly operations. The project will study agent-based methods in order to turn traditional, hard-wired operations into modern, adaptive ones capable of coping with the requirements of flexibility and parts uncertainty. Agents in the context of this proposal are sensing, actuating and reasoning/controlling devices with behaviours that can also be modelled as agents (see Fig. B1.9).
An agent-based approach has many advantages. Its modular structure allows reuse of agents for different tasks and applications. Because an agent-based approach is decentralised, short response times can be achieved providing real-time capability. Independence of agents leads to an increased robustness of the overall system in the face of failure locally. In a distributed implementation of multiple agents, self-adaptation can be enhanced through inter-agent communications relaying environment changes to other agents. Building on the proposed agent-based framework, agents can be created to provide self-adaptive solutions for assembly, testing and inspection. It is expected that a number of agents (such as, for example, force and torque sensor agents and actuation agents) can be reused across different operations.
Self-adaptivity – Specific application scenarios
The proposed agent-based concept will be applied to the three specific scenarios discussed above, Hub-bearing assembly, Yaw testing, Upper housing inspection. Agents will be formulated based on the general “sense-reasoning&control-acting model” and the specific requirements for each case (see Fig. B1.9 Table B1.1 and the Requirement analysis section). Self-adaptivity is achieved by providing each agent with an appropriate cost function and error metrics for the solution of the task; an optimisation step will enable the agent to complete a task successfully despite environment changes and uncertainty. Where needed, closed-loop hybrid (force and position) feedback control using multiple sensor inputs from vision, force and torque sensors will be employed. Adaptivity in vision will be achieved through intelligent methods recognising image segments that constitute specific, task-dependent features despite uncertainty in position and noise in camera images; input from CAD will be incorporate where possible. Integrating sensor signal input (such as force/torque sensors) and material/environment knowledge with model-based approaches, deviations from modelled behaviour will be predicted and used to feed into the adaptive control strategy. The identification/optimisation of model parameters can be achieved through a training procedure based on Artificial Intelligence (AI) approaches will be investigated in this context. Learning will be achieved using appropriate training algorithms and features extracted from experimental data as well as input from experienced users. The generalisation capabilities of these AI-based methods will provide the required adaptivity which has the potential to be further modified through on-line learning (self-adaptivity). The automated on-line interpretation of sensor signals will enable real-time error detection and error recovery.
PROCESS VARIABLES / PARAMETERS SELF-ADAPTIVE STRATEGIES
Assembly
(Hub-bearing Assembly) Force/time
Bolt elongation
Bolt material properties
Bolt position
Lack of parallelismOvalityCAD models - Adaptive control including Non-linear control, Neural network / Fuzzy logic based control, Model-reference and -predicive control
- Intelligent image recognition and analysis tools
- Optimisation techniques including Newton Raphson, Least square
- Learning methods (supervised, unsupervised)
Functional Testing
(Yaw testing) Yaw friction
Yaw vibration
Radial clearance
Hydraulic escapes - Emphasis on sensing, less on control/actuation
- Model-based methods for prediction of dynamic yaw behaviour
- Predict in field performance
- Fuzzy logic to model vague, human expert knowledge
Inspection
(Upper housing inspection) Vision
Surface laser profiling (3D)
Distances (ultrasonic sens) - Emphasis on sensing, less on control/actuation
- Sensor fussion
- Intelligent image recognition and analysis tools
- Intelligent adaptation to variability in product dimensions
Project Results:
Market (and stakeholder) analysis
This chapter contains a market analysis of the segments related to the exploitable results derived from COSMOS. Only select markets are analysed this way as some of the results have overlapping segments and some results are academic and as such usable in further research and not as a product.
This analysis explores the possible chances of products or services derived from COSMOS results on the open market and are used as a basis for further exploitation plans.
2.1 MES
MES is emerging as a superior product for flexible and integrated production. An MES combines all the properties required for modern production management in a comprehensive software system. This system is the basic and operational platform for all the enterprises evolving towards the digital factory of the future. MES solutions are increasingly used and recognized in the process manufacturing and discrete manufacturing industries (automotive, healthcare, aerospace and defence, FMCG etc.). This trend is accelerated by the ongoing economic globalization that requires maximized competitiveness.
The Global Manufacturing Execution System market will have a growth rate of 11.7 percent over the period 2011-2015. One of the key factors contributing to this market growth is the increasing demand from discrete manufacturing industries. Further, the Global Manufacturing Execution System market has also been witnessing emerging demand for MES solutions for mobile computing devices.
North America and Europe account for similar market share in the global MES market. Since the APAC region is becoming the manufacturing hub for many industries, this region possesses great potential for the MES market and most of the major vendors are focusing APAC now. Clearly, there are many vendors that offer MES software, but a big chance still exists out of question because of the potentiality and expandability of this market.
In the field of MES we differentiate between three topics: factory control, automatic manufacturing system and interoperability middleware, which gets its own chapter because of its importance.
2.2 Robotics and Automation
The situation concerning the production sector of wind turbines towards the end of the COSMOS project is assessed by Reis Robotics as robot manufacturer and system integrator as follows:
In the current period the investment rate in the wind turbine sector is rather low. Especially in Germany, this is mainly related to the existing over capacity of the amount of current being produced by the installed wind turbine plants. This over capacity is mainly resulting from a lack of power lines to convey the produced current from the production sites to the main areas of the power consumption. In many cases, investors for wind turbine installations postpone their activities until there will be a solution of the current transportation problem. This situation is especially valid for Germany and may be similar in other European countries. The situation all over Europe, however, seems to be heterogeneous.
Currently it can be stated from the Reis point of knowledge that the amount of realised wind turbine systems is rather reduced than expanded. Also the actual construction activities of new wind turbine plants are rather low, due to the described situation.
The COSMOS results are in the context of the project exemplarily demonstrated in the specific production step “bolt tensioning at the rotor hub of a wind turbine”. Principally, the self-adaptive production principles could also be applied to other production steps of the wind turbine production. Even other production scenarios in the component production, predominantly in the production of components in the renewable energy sector, could be addressed by the COSMOS results, as e.g. the battery production that becomes important in relation to this sector in future. A further innovative sector that can be addressed by the results could be the e-mobility area, especially in the production of lightweight parts for e-cars.
The sketched facts show that in any case there are several possibilities for a sustainable exploitation of the COSMOS principles of self-adaptiveness in practice in future times.
2.3 Interoperability Middleware
2.3.1 Communication interfaces of devices
Standard automation devices like PLCs offer different communication interfaces today. One group is field busses mainly used to connect I/O modules to the PLC but they are also used to connect the PLC to process management level. Another group is vendor specific TCP/IP based protocols. Both groups either communicate byte blobs without any type information or a list of variable values. OPC Data Access servers are available for all standard automation devices. They encapsulate the proprietary communication protocols in a standard interface on Windows PCs. OPC Data Access servers also limit the interface to a list of variables with values.
Special devices provide either serial connections or proprietary TCP/IP protocols. They often offer libraries for application programmers on Windows PCs.
MES protocols cannot be implemented on embedded devices since they are product specific and resource consuming. In addition most embedded device vendors do not allow to put any external components on the device.
2.3.2 Communication interfaces of MES systems
MES servers typically provide a vendor specific interface with a list of services. The communication is done through Enterprise Service Bus (ESB), web services, CORBA or vendor specific HTTP based protocols. Some vendors provide libraries for application programmers on Windows PCs.
Device protocols can normally not be implemented on a MES server since they communicate through the separation of process and IT networks.
2.3.3 Today’s integration strategy
The communication integration between devices and MES is typically done on line controller PCs or communication PCs in the production environment. On these PCs standard software of the MES is used to map supported protocols like OPC or Modbus TCP to the services and communication protocols used by the MES. If the device protocol is not supported, project specific software is developed that maps the device specific protocol to the MES services.
2.3.4 Status OPC UA integration on device level
OPC UA server functionality will be standard on PLCs. It is already supported by several vendors and much more have support announced. A major step will be support by the biggest vendors Siemens and Rockwell. Support for OPC UA client functionality was announced by several vendors and is important for vendors to allow device to device communication. The full functionality required by COSMOS is not supported yet.
Several machine builders have added OPC UA as standard interface to their products. Device vendors see OPC UA as the fallback communication channels to PLCs and other systems if the supported field bus is not used in the project. Prototypes exist but not many products.
The first step would be to integrate the necessary OPC UA functionality required by COSMOS OPC UA device service bus in a relevant number of devices. This requires updates of the device firmware. Standardized plug-and-use functionality can be added on a configuration level in most devices.
2.3.5 Status OPC UA integration on MES level
MES vendors like SAP and iTAC have integrated OPC UA into their products already. The level of support does not cover all aspects of the COSMOS OPC UA device service bus but there is a very high interest of MES vendors to standardize integration with the process layer. Therefore the concept of the COSMOS OPC UA device service bus is very interesting for MES vendors. Level of support will depend on level of support in devices. The first step is to integrate all necessary OPC UA features into MES systems.
The second step is the further standardization of the semantic of plug-and-use functionality. This will further simplify integration of process data with MES.
3 Positioning of COSMOS in standardization bodies
Standardization is one of the key success factors for COSMOS to reduce integration effort and costs for devices, subsystems and modular factory units. Therefore ASCOLAB is active in several different standardization working groups that focus on standardizing integration of different systems from the device level up to the enterprise level.
3.1 OPC Unified Architecture Working Group
The OPC Unified Architecture (UA) working group is a working group of the OPC Foundation. This is a permanent working group that defines the basic OPC UA functionality in a multipart specification with currently 13 parts. The results of this working group are also distributed through a liaison with IEC as international standard IEC 62541. Parts 1 – 10 are already available as IEC versions.
ASCOLAB is an active member of the working group since its start in October 2003. ASCOLAB is as editor responsible for two parts of the specification, Part 4 – Services and Part 9 – Alarms and Conditions.
OPC UA specification version 1.02
During the COSMOS project the OPC UA working group worked on a new release of the OPC UA specification with version number 1.02. Most of the updates released with this version in August 2012 were mainly maintenance enhancements and clarifications. Only a few new features were added to the existing parts. Two new parts were released in this version, Part 11 – Historical Access and Part 13 – Aggregates.
OPC UA Part 12 - Discovery
Another new specification, Part 12 – Discovery is relevant for COSMOS and defines enhanced discovery functionality necessary for the OPC UA device service bus in COSMOS. This specification part is not released yet since it requires prototype implementation and detailed evaluation. A first release candidate for final review and voting will be available in December 2013.
The two relevant features for COSMOS OPC UA device service bus in OPC UA Part 12 are the ad-hoc service discovery and the Global Directory Service (GDS).
Ad-hoc service discovery allows OPC UA servers to announce their service functionality in a local subnet using mDNS protocol. OPC UA clients can find OPC UA servers and their service capabilities by accessing a local mDNS protocol cache in the Local Discovery Server (LDS). ASCOLAB implemented two prototypes of the mDNS protocol integration into the LDS. This multicast extension to the LDS is used by servers to announce their functionality in the network and by clients to access the local mDNS cache. The second prototype will be the base for the new LDS-ME (Local Discovery Server with Multicast Extension) provided by the OPC Foundation as standard discovery component.
GDS is a standardized OPC UA interface that encapsulates a system for plant wide OPC UA application configuration and discovery. The discovery functionality extends the subnet discovery through mDNS protocol and provides a global service registry that works plant or corporate wide across different subnets. The second part of the interface allows a central configuration of application level security by managing and distributing security certificates and application trust relations. The interaction between the GDS and OPC UA applications is defined in Part 12. The internal implementation of a GDS can be a wrapper around an existing system like a Windows Active Directory or a stand-alone application for central OPC UA device service bus configuration. ASCOLAB developed a GDS prototype for evaluation in the OPC UA working group.
3.2 OPC Foundation and PLCopen joined working groups
PLCopen is the organization that maintains the international standard IEC 61131-3. This standard defines PLC programming languages and a software model for PLCs. The joined working groups between OPC Foundation and PLCopen define essential functionality necessary to fully integrate off the shelf PLCs into the OPC UA device service bus.
OPC UA Information Model for IEC 61131-3 (PLCopen)
This working group defines the details for integration of an OPC UA server into PLCs. ASCOLAB is an active member of the working group as editor of the specification and joined all working group meetings. The first version of the OPC UA Information Model for IEC 61131-3 defines a mapping of the IEC 61131-3 software model to OPC UA objects and covers mainly the data access use case. The first version is released since March 2010
The mapping of additional OPC UA features like Methods and Events is currently under definition in version 2.0. These features are required for integrating PLCs like the one from PHOENIX CONTACT with all features into the OPC UA device service bus for COSMOS. ASCOLAB presented use cases to the working group.
OPC UA communication function blocks for IEC 61131-3
This working group defines details for integration of an OPC UA client into PLCs. ASCOLAB is an active member of the working group as OPC UA expert and joined all working group meetings.
The working group defines function blocks that allows a PLC programmer to act as OPC UA client from within the PLC program. All OPC UA client functionality necessary for COSMOS is covered by the first specification version. The functionality is an essential part for the integration of PLCs into the OPC UA device service bus.
The first results of the working group with a working prototype between a Beckhoff PLC acting as OPC UA client and the MES system from iTAC acting as OPC UA server were shown at the Hannover Fair in spring 2013. The two PLC vendors Beckhoff and B&R Automation finished a complete prototype implementation of the defined function blocks in September 2013. Based on the prototyping results a release candidate specification is planned for December 2013.
3.3 MES Connectivity
This OPC Foundation working group was started in December 2011 together with PLCopen, MESA (Manufacturing Enterprise Solutions Association), VDMA (German Machine Builder Association), ZVEI (German Electro Technical Industry Association) to standardize integration of devices and MES based on OPC UA. The requirements for the OPC UA device service bus in COSMOS are completely covered by the use cases defined by this working group.
ASCOLAB is one of the initiators of the working group and active as editor of the working group.
3.4 OPC DI version 2.0 and FDI (Field Device Integration)
Field Device Integration (FDI)
The FDI working group defines an international standard for configuration and maintenance of field devices. The working group consists of the standardization groups PNO (Profibus User Organisation), HART Foundation, OPC Foundation, FDT Group (Field Device Tool), Fieldbus Foundation and EDDL (Electronic Device Description Language). FDI will integrate and harmonize all of these standards and will integrate them into one common standard.
ASCOLAB is member of the review team and the handheld working group for the new FDI standard.
OPC UA for Devices (DI)
OPC DI is the base OPC UA information model for FDI. Generic features for device configuration defined by the FDI working group were integrated into OPC DI version 2.0. This new specification version was released in July 2013.
ASCOLAB provides the chairman for this working group.
3.5 OPC UA Information Model for ISA95
The goal of this OPC Foundation working group is the definition of a mapping of the established standard ISA 95 to OPC UA. The working group was started in January 2012.
HOLOS and ASCOLAB are represented in the ISA95/OPC-UA Working Group. The goal is to develop the standard with the contribution of Cosmos requirements and use cases. This will enable the use of the standard in COSMOS. Using this standard in COSMOS enables the plug and use functionalities and integrates in COSMOS a standardized communication interface. This standard is already used worldwide through B2MML.
The specification was distributed as release candidate for voting to the OPC Foundation Technical Advisory Council (TAC) in October 2013. ASCOLAB as member of the TAC and a majority of members voted for the release of the specification. The distribution of the final release document is expected for November 2013.
3.6 AutomationML, IEC SC 65 E, Project no. 62714 Ed 1.0
The goal of this working group is to standardize engineering data exchange for use in industrial automation engineering.
EDAG is a Steering Committee member of the AutomationML group and active in the working group. EDAG has evaluated AutomationML for use in the COSMOS project.
4 Business model and COSMOS exploitation strategy
In this chapter business models will be discussed in greater detail. This includes existing business models of some partners involved in COSMOS, which are showcased here as an example. Additionally a more general approach of possible business models for exploiting the results of COSMOS is illustrated.
4.1 Existing business models of partners
In this chapter the business models of two COSMOS partners will be showcased. To describe these business models the CANVAS schema is used. For the sake of simplicity the usual CANVAS panel is transformed into an easily readable table.
4 1 CANVAS explanation
4.1.1 EDAG
Key partner - COPADATA
- FFT
- Employees
Key activities - Consulting for IT systems
- Problem support for IT systems
- Development of concepts for IT systems
- Development of IT systems
Key resources - Developers, project managers, consultants
- Development environments
Value proposition/services - Process improvements
- Automating through IT
- Analysis and reporting
- Process development
- Software development
Relationships - Technology partner: Support for partner applications, development using partner solutions
- Affiliated company: Temporary distribution of manpower, joint processing of inquiries
- Key user: improvement ideas, new inquiries
Channels - Inquiries from OEMs: distributed from company and affiliates
- Key-accountant: advertising services to OEMs
- Follow-up orders from regular customers
- Trade shows (both internally and externally organized)
Clients, customer segments - OEMs: consulting, support and development MES and ERP systems
- Segments: automotive, aerospace, rail, general assembly, medical and component suppliers
Costs, investments - Team: costs for team members (personnel, travel, facilities)
- Time: time employees and managers spend within workshops/projects.
- Development tools: license costs, hardware costs, administration costs.
- External consultants: subcontracting for specialized expertises
Revenue improvements, cost savings wins - Pricing: dependant on time invested (personnel and travel expenses included)
- Value: process improvements reduce running costs by reducing needed resources (time, personnel, equipment), quality is improved and
4.1.2 IBERMATICA
In this subchapter the business model of IBERMATICA is described using the CANVAS-schema in tabular form.
Key partner - PARTNERS: Microsoft, IK4 – TEKNIKER
- Quality -> CMMI
- Providers: PENTA, SIEMENS, ROCKWELL, MATRIKON
- ERP providers: SAP, AXAPTA, NAVISION, RPS, ...
- Employees! Key driver/initiators for improvement
Key activities - Process improvement consultancy/service
- Problem solving support
- Process monitoring, decision system support, taceabilty
Key resources - Project managers, MES consultants, quality consultants
- MES modelling software, data mining
Value proposition/services - Real estate plant monitoring
- Presence control, work oder tracking, real downtime control
- Management, documentation, production support
- Maintenance management
- KPI analysis and reports
- Support for continuous improvement process
Relationships - Process council: Meetings of top management with process owners and managers in order to improve processes
- IT & Quality: Support from IT for process automation. Support quality to achieve ISO requirements.
- Technology partner: Customer support in the application of new technologies to manage production
Channels - Cross selling: ERP Solutions, CAM, quality, maintenance
- Process improvement projects: Strategic projects to redesign/heavily improve processes
- Communication via multiple channels: Intranet, workshops, product presentations
Clients, customer segments - Upper management: Provide tools to improve process performance.
- Middle management: Driving operational effiency (more with less).
- Employees: Making their daily job easier by providing a system to capture process data.
- Segments: Automotive, aerospace, food, plastic injection, manual assembly.
Costs, investments - MES Team: Costs for MES team members (personnel, travel, facilities)
- Time: Time employees and managers spend within workshops/projects.
- OPC, .NET tools: License costs, hardware costs, administration costs.
- MES consultants: Bringing in external consultants to support with specific expertise.
Revenue improvements, cost savings wins - Pricing: Free of charge for internal improvements. Consultancy fee for external improvements.
- Value: Increased revenue from improved processes. Reduced lead time. Reduced costs by improving quality & speed.
4.2 Possible business models
This chapter explains different general business models relevant to the COSMOS partners. While not all models are applicable to every result, these present possible approaches to facilitate the exploitation of the results.
4.2.1 Service-oriented
A service-oriented business model concentrates on delivering a certain service, for example consulting, usually for a hourly or daily fee. In this model, the knowledge and experience of the employees is the greatest asset and revenue is generated by using this asset to improve the processes of the client.
Key partner - Consultants: provide knowledge and experience
Key activities - Consulting
- Support
Key resources - Employees
- Knowledge
- Experience
Value proposition/services - Consulting improves processes
- Support to find errors and minimize outages
Relationships - Technology partners: Support for certain technologies used
Channels - Trade fairs
- Technology partners
Clients, customer segments - Businesses using the tools/processes the employees are experienced with
Costs, investments - Employees
- Equipment and tools for the employees
Revenue improvements, cost savings wins - Time based fees for consultation and support.
4.2.2 Solution provider
A Solution provider is contracted to find a solution to a certain problem. The solution itself can be a process, a tool or a concept to counteract the problem. Revenue is generated by selling the solution, usually at a price determined at the beginning.
Key partner - Consultants: provide knowledge and experience
Key activities - Development of tools/processes/concepts
Key resources - Employees
- Knowledge
- Experience
Value proposition/services - The problem is solved for the client or an existing process was optimized and is now faster/cheaper/improved
Relationships - Technology partners: Support for certain technologies used
Channels - Trade fairs
- Technology partners
Clients, customer segments - Businesses wanting to solve a certain problem hindering their productivity
Costs, investments - Employees
- Equipment and tools for the employees
Revenue improvements, cost savings wins - Contract fee for the solution as a whole
4.2.3 Component provider
A component provider supplies a finished product to customers in need of a certain tool or resource. Supporting and marketing this product is usually also part of this business model. Revenue is generated by selling units of the product or licenses allowing its use.
Key partner - Marketing partners
- Distributors
- Possibly subcontractors
Key activities - Marketing and supporting the product
- Further development of the product
Key resources - Employees
- Experience
- Resources needed to create the product
Value proposition/services - The product itself as a tool or component needed to create another product or a service
Relationships - Technology partners: Support for certain technologies used
- Marketing: Generating a market for the product
- Customers: Feedback for product improvements
Channels - Trade fairs
- Technology partners
- Trade with contractors and/or subcontractors
Clients, customer segments - Customers in need of a certain tool or resource fulfilled by the product
Costs, investments - Employees
- Equipment and tools for the employees
- Resources to create the product
- Further development of the product
Revenue improvements, cost savings wins - The product as a tool or resource to generate revenue or fulfil a certain need
4.3 COSMOS general exploitation strategy
This chapter contains information about the general strategy used as a basis for the detailed exploitation strategies. The possible business models serve as a context in which the results can be exploited.
The final results of the COSMOS project depicted in the figure below are referenced by the table following it. This table gives an overview over the exploitation plans detailed later in the document and states the business model(s) to be associated with every exploitation strategy.
4 2 Overview COSMOS final results
N° Exploitation strategy Final result Relevant business model Owner
1 Automation ML Connector for MES (AML2MES Adapter)
Automation ML Connector Component provider and service-oriented EDAG
2 Registration and CAD2Robot Trajectory Planning
Robot Based Inspection Solution provider KCL
3 Self-adaptive Bolt Tightening
Bearing Assembly Solution provider KCL
4 Self-adaptive Testing
Computer Based Texting Solution provider TEKNIKER
5 Self-adaptive Assembly System
Bearing Assembly Solution provider REIS
6 Factory Control/SCFU Control
Factory Control Solution provider IBERMATICA
7 Interoperability Middleware
Interoperability Middleware Component provider and service-oriented ASCOLAB
Potential Impact:
In this section the exploitable results of the Cosmos project will be briefly described, also highlighting the main progress beyond the State of the Art. The main “owner” for each Cosmos component will be identified and the owner/s will be the main responsible for the exploitation of the specific part of the Cosmos project.
The following table reports on exploitable results of the project which the partners intend to exploit.
N° Result Has the result commercial/ social significance? Can the result be exploited as a stand alone product, process, service, etc? Should the result be grouped with one or more others? Should the result be split into one or more others? Reporter
1.0 Factory Control IBERMATICA
1.1 Factory control: Lean production support Yes No Yes No
1.2 SCFU workflow definition Yes No Yes No
1.3 SCFU control. MES components Yes No Yes No
1.3.1 OPC/UA Ibermatica Yes No Yes No
1.3.2 OPC/UA EDAG Yes No Yes No
1.3.3 Automation ML adapter EDAG Yes No Yes No
1.4 SCFU control: HMI content editor Yes No Yes No
2.0 Flexible automatic manufacturing system REIS ROBOTICS
2.1 Self-adaptive control algorithms Yes Yes No Yes KCL
2.2 (Self-adaptive) geometrical verification system Yes Yes Yes No stt, KCL
2.3 Self-adaptive assembly process Yes Yes Yes No ZEMA, REIS ROBOTICS, KCL
2.4 Self-adaptive inspection process Yes Yes No No
2.5 Economic model and evaluation tool Yes Yes Yes No RWTH WZL
3.0 Interoperability Middleware
3.1 Automation ML connector for MES Yes Yes No Yes EDAG
3.2 OPC/UA connector for Phoenix PLC Yes Yes No Yes ASCOLAB
3.3 OPC/UA connector for MES Yes Yes No Yes ASCOLAB/ IBERMATICA/ EDAG
3.4 B2MML-based connector for MES/ERP Yes Yes No Yes IBERMATICA
3.5 Knowledge on plug and use functions/standardisation Yes Yes No Yes ASCOLAB/ HOLOS
3.6 OPC/UA connector for OPC data access Yes Yes No Yes ASCOLAB
Table 3 1 Exploitation results
3.1 Market Analysis
This chapter contains a market analysis on three major topics COSMOS is dealing with.
The analysis contains a description of the innovations the products brings, an outline of potential customers and what benefits the customers may expect. In addition to that, dates for the estimated timeline to market the product are given and costs for the exploitation are summarised. Furthermore the potential product price range and market volume are estimated. Then, a list of competitors is compiled and an explanation on how the new product will rank against their products and how fast they are expected to react. Finally all involved partners are listed and the modalities of protecting the product are named.
This chapter contains a market analysis on three major topics that COSMOS is dealing with. All the three parts are based on Self-Contained Factory Unit (SCFU) and are important elements of a Manufacturing Execution System (MES).MES is emerging as a superior product for flexible and integrated production. An MES combines all the properties required for modern production management in a comprehensive software system. This system is the basic and operational platform for all the enterprises evolving towards the digital factory of the future. MES solutions are increasingly used and recognized in the process manufacturing and discrete manufacturing industries (automotive, healthcare, aerospace and defense, FMCG etc.). This trend is accelerated by the ongoing economic globalization that requires maximized competitiveness.
The Global Manufacturing Execution System market will have a growth rate of 11.7 percent over the period 2011-2015. One of the key factors contributing to this market growth is the increasing demand from discrete manufacturing industries. Further, the Global Manufacturing Execution System market has also been witnessing emerging demand for MES solutions for mobile computing devices.
North America and Europe account for similar market share in the global MES market. Since the APAC region is becoming the manufacturing hub for many industries, this region possesses great potential for the MES market and most of the major vendors are focusing APAC now. Clearly, there are many vendors that offer MES software, but a big chance still exists out of question because of the potentiality and expandability of this market.
We differentiate the MES here in Factory control, Flexible automatic manufacturing system and interoperability middleware three topics.
Factory control
Describe the innovation content of result 1 Distributed Control Architecture in Assembly Processes, Factory and SCFU control = SW packages
Who will be the customer? Industry
What benefit will it bring to the customers? Modular FActroy Control System with new Standard communication links (OPC UA) and For- Backward data exchange with engineering systems.
Modular Manufacturing Execution System with Interoperability with ERP system.
Ease of handling by non expert users. Components for HMI creation and workflow definition
When is the expected date of achievement in the project (Mth/yr)? Aug 2013
When is the time to market (Mth/yr)? Jun 2014
What are the costs to be incurred after the project and before exploitation? 200 K€ (EDAG)
275 K€ (IBER)
What is the approximate price range of this result / price of licences? MES solution 25 k€
OPC UA 3k€
AutomationML 35 k€
What is the market size in M€ for this result and relevant trend? 25 M€/a (IBER) + 10 M€/a (EDAG)
How this result will rank against competing products in terms of price / performance? Us achieving this goal by developing a tool that allows end users to easily configure their own solution.
OPC UA – same price/performance
AutomationML no competing product
Who are the competitors for this result? (Fraunhofer IOSB), Siemens (EDAG)
SYSTEPLAN, IDS, MAPEX (IBER)
How fast and in what ways will the competition respond to this result? 3 to 5 Years
Who are the partners involved in the result? OPC UA: Ascolab,
Workflow definition : TEKNIKER
AutomationML (none)
Who are the industrial partners interested in the result (partners, sponsors, etc…)? Ibermatica, EDAG, Phoenix, Gamesa
Have you protected or will you protect this result? How? When? By the end of the project.
Table 3 2 Factory Control exploitation results
3.1.2 Flexible automatic manufacturing system
Describe the innovation content of result1 Result is a piece of machinery / a process / a method; within the context of a wind turbine production line; flexible production method (self-adaptive); new production method; new technology; strength: improved flexible production process
Who will be the customer? Customers from the wind turbine sector and/or also from other suitable industries.
What benefit will it bring to the customers? More flexible production process.
When is the expected date of achievement in the project (Mth/yr)? 05/2013 (approx. 3 months before project end).
When is the time to market (Mth/yr)? Project end + 2 years: 11/2015.
What are the costs to be incurred after the project and before exploitation? 6-12 person months (only rough guess).
What is the approximate price range of this result / price of licences? 100,000 Euros.
What is the market size in M€ for this result and relevant trend? 5 systems per year (5 x 100,000 Euros).
How this result will rank against competing products in terms of price / performance? Price: higher (20%); performance: better (30%).
Who are the competitors for this result? Other system integrators in the wind turbine production line sector.
How fast and in what ways will the competition respond to this result? 2–3 years after launch of COSMOS result.
Who are the partners involved in the result? Reis, STT, Phoenix, WZL, KCL, ZeMA
Who are the industrial partners interested in the result (partners, sponsors, etc…)? Reis, STT, Phoenix
Have you protected or will you protect this result? How? When? If possible and necessary: by patents. Currently open, whether the COSMOS results can/will also be patented.
Table 3 3 Flexible automatic manufacturing system exploitation results
3.1.3 Interoperability Middleware
Describe the innovation content of result1 Knowledge in operation and development of Interoperability Issues, integrating OPC/UA
Who will be the customer? Industry, Self.
What benefit will it bring to the customers? Pluggability infrastructure for different component types (software and hardware)
When is the expected date of achievement in the project (Mth/yr)? 1st Year
When is the time to market (Mth/yr)? Two years.
What are the costs to be incurred after the project and before exploitation? ~50 K€
What is the approximate price range of this result / price of licences? ~ 100 K€
What is the market size in M€ for this result and relevant trend? 3 M€
How this result will rank against competing products in terms of price / performance? Better, cheaper and faster regarding implementation costs.
Who are the competitors for this result? All companies implementing DPWS and OPC/UA.
How fast and in what ways will the competition respond to this result? 1 to 2 Years
Who are the partners involved in the result? Ibermatica, Tekniker, Ascolab, Holos
Who are the industrial partners interested in the result (partners, sponsors, etc…)? Ascolab, Reis, Holos, Edag, Phoenix
Have you protected or will you protect this result? How? When? By the end of the project.
Table 3 4 Interoperability middleware exploitation results
Various dissemination tools that have been developed by the project so far and outlines their purpose as means of sharing information about the project and generating interest in the work being undertaken by the COSMOS project.
3.1 COSMOS website
Target audience: Public, interested 3rd parties
A website has been created for the project and can be accessed using the following url: http://www.cosmosproject.eu/
As the project’s main dissemination channel, the website serves as a source of information about the project for the public and all external parties who may have an interest in the project. The maintenance and updating of the COSMOS website is done by IBERMATICA, project coordinator. However, the input for the content is generated by all of the project partners.
The website provides the following information:
• General static information about the project, including the project concept, partners, volume and contact information.
• News, events and press releases
• Publications; public deliverables and other publications as and when they are available
3.2 Dissemination material
3.2.1 Project Presentation & Leaflet
A Leaflet was prepared and publish with a brief project presentation in English, containing key information about the project (participants, approach, expected achievements etc.). In addition, this presentation was attached also on the project’s website and within other dissemination activities.
Figure 1 - Cosmos Brouchure A
Figure 2 - Cosmos Brouchure B
3.2.2 Roll-up posters
For a more effective means of visual communication and dissemination several roll-ups were designed and printed to be exposed in conferences and industrial fares and set up alongside the project exhibition booth.
3.2.3 Videos
Several videos were created during the project in order to expose advances.
3.2.3.1 Self adpative tensioning process at the rotor hub
Within the Cosmos Project a self-adaptive tensioning process and tool are developed which allow an exact preload application to the bolts of a wind turbine pitch bearing. In the first video the tensioning process is shown at the demonstrator which consists of six bolts and is used for testing the process and the tool. Moreover, the video presents the human interface to control the process and monitor different parameters. In the first step of the process the tensioning tool stops over the bolt. In the next step the tool is moved linear onto the bolt till it gets into contact with the surface of the bearing ring. If the contact is detected the change bushing is threaded onto the bolt and pressure is built up. In this condition the bolt is stretched and the nut is tightened. At the End of the process the pressure is released and the tensioning cylinder removed.
http://www.youtube.com/watch?v=LdGqWRqkcnU&feature=youtu.be
Figure 3 - Self adaptive tensioning process at the rotor hub
3.2.3.2 Self adaptive tensioning process at the demonstrator
The results of the initial tests of the demonstrator a used for the validation at real hub of a 2MW wind turbine. This video shows the first tests of the tensioning process at the real rotor hub. At the beginning the position of the bearing is detected. Therefore, the robot measures different points of the rotor hub with a laser sensor. The tensioning points of the bolts are calculated within the measured rotor hub coordinate system with the product parameters. So the tool is guided by the robot to the different tensioning points where the bolts are tensioned one by one.
http://www.youtube.com/watch?v=4WZGgnrr2PE&feature=youtu.be
Figure 4 - Self adaptive tensioning process at the demonstrator
3.2.3.3 COSMOS Geometrical Inspection
Self adaptive testing and inspection. The object inspection is performed with 3D laser scanner fixed on industrial robot. The surfaces are scanning according to CTQs (Critical To Quality) definition. The robot trajectory planning are originates from object CAD drawing and CTQs.
http://www.youtube.com/watch?v=IivHjThjm2g&feature=youtu.be
Figure 5 - COSMOS Geometrical Inspection
List of Websites:
www.cosmosprojec.eu