Service Communautaire d'Information sur la Recherche et le Développement - CORDIS

Final Report Summary - FACTORY-ECOMATION (Factory ECO-friendly and energy efficient technologies and adaptive autoMATION solutions)

Executive Summary:
The energy costs represents a small part of the production expenses, meaning that the design of production facilities usually neglected investments in efficient energy usage in favour of more important factors, such as the maximization of production throughput. The energy consumption of factories has been in fact always high, but the cost of energy has increased dramatically in the last decades, pushing the industrial representatives to search for higher level of efficiency This scenario, however, is changing rapidly. According to the International Energy Agency, manufacturing is responsible for approximately 37% of global primary energy consumption being the largest energy consumer and CO2 producer. Improvement of energy/resource efficiency is the key to reduce environmental impact and is mentioned as driver to reach European 20/20/20 goals.
In order to obtain an efficient work and material flow, and achieve such challenging results, an overall optimization of the whole concept of Factory has to be pursued, introducing a new holistic vision of the production environment, with comprehensive interventions acting both at methodological and technological level.
The aim of the Factory-ECOMATION project has been to enabling European manufacturing industries to overachieve Europe 2020 program targets through development of breakthrough innovations for cost-effective, highly productive, energy-efficient and near-zero-emissions production systems, by means of:

• Definition of a holistic perspective of the economically and ecologically oriented production environment;
• Development of a comprehensive sensing, monitoring and data evaluation system which could grant access to an organic view of all the materials, energy, wastes and emissions flows within the factory;
• Extension and development of new technologies to increase energy efficiency by reducing its consumption at machinery and production level, and by recovering it whenever wasted;
• Development of advanced emissions abatement technologies
• Introduction of a new adaptive management and automation platform to optimize production taking into account not only productivity targets but also eco and energy-oriented ones.
To assure the impact on European economy, Factory-ECOMATION partnership includes relevant Industrial players in technology development and manufacturing, together with highly-recognized Research Centres. The inclusion of IMS partners will also grant a global scope to the project objectives, essential to achieve all the proposed results, including standardization and definition of eco-labelled processes.
The project results has led to a sustainable, as well as economically profitable, energy efficient, near-zero emissions factory framework, based on new measurement and control technologies and on new adaptive factory automation solutions. The new established paradigm is intended to become a permanent reference point in sustainable European Manufacturing.

Project Context and Objectives:
The strategic key for European Manufacturing “Factories of the Future” is to invest in the production of high-quality technical products and/or services welcoming transformation of the manufacturing industry into a high added value and knowledge-based industry, which is competitive in a globalised world. As a result, the manufacturing vision needs to be renovated and enriched: the production environment needs to be deeply evolved compared to existing reality, also in everyone system conceptual frame.
Nowadays, there is a concentrated effort in scientific world to find alternative sources of energy. Emphasis is on renewable energy like wind, solar, hydrogen, etc. To encourage further research into the alternative sources of energy, there is an increased pressure for enforcing pollution taxes and in particular carbon tax. [Baranzani et al., 2000] presented the advantages of applying carbon tax while [Painuly, 2001] proposed the usefulness of green credits to encourage the adoption of renewable energy sources.
However, it is also a reality that these alternative energy sources are not available in immediate future: fossil fuels are not only the primary source of energy today but energy projections of Environmental Information Agency show that even in the year 2030, this situation is not going to change. For these reasons, along with finding alternative energy sources, efforts must be made in the industrial sector to seek modus operandi that could, on the short to medium term, minimize the damage caused by the use fossil fuels. Within this context, one of the most important solutions is the promotion of a more rational use of energy or, in other words, increasing the energy efficiency of industrial processes.
During the last decades, pushed by eco-oriented policies and standards, the European manufacturing industries has seen a clear path of introduction of new technologies within the production environment (or on its boundaries) conceived to reduce energy consumption, recycle wastes or filter emissions. However, in the majority of those cases, “production is still the king” and these systems are regarded mainly as a support function whose objective is to provide additional services to the production unit. Within this situation, in spite of the fact that the activity level of these add-in systems is dependent on the production management of the main process, the energy, waste and emissions considerations are rarely included during design phase or in the planning, scheduling and control procedures.
It is under this context that Factory-ECOMATION main high-level objective is to conceive, define, design and demonstrate, through adequate industrial pilots, a new paradigm of factory where energy, emissions, and environmental-related aspects in general, are considered and included into the structure and management of the factory itself, at the same level and with the same importance of traditional criteria such as productivity and products quality.
Our idea of “Factory of the Future” is based on a concept of complete integration of its constituting elements: production machines, energy recovery systems, emissions abatement equipments, and the corresponding flows on which they act, be they related to energy, materials, wastes, finished products or emissions. The reason behind this approach is to jump beyond the limits imposed by each single technological improvement taken on its own, by leveraging and exploiting the potential synergies guaranteed only through their coordinated interactions.
Within this context, Factory-ECOMATION will introduce a key concept influencing the entire project, the Unified Flows approach. In order to achieve the results expected, in terms of reduction of the environmental impact of the factory, it is necessary to change the whole idea of factory, which shall not be considered anymore as just an entity to create products through the processing of materials and energy, for which wastes and emissions are an unconsidered side-effect. In Factory-ECOMATION factory each input and output flow will be put on the same level, transforming, at the same time, its high-level targets; in practical terms, throughout the entire design, monitoring, analysis, management and control of the factory, the input-energy, materials, output-products, wastes, wasted-energy and emissions flows will be included under a unified vision, in which potential transformation among them will be analyzed and considered for the computation of the real performance of the factory. This could mean, taking this vision to its extreme consequences, that a production of wastes could be seen as positive and eventually even increased (by modifying the production schedule) because those wastes can be transformed in energy (through specifically developed energy recovery technology), that in turn could be used to increase efficiency and contemporarily power an emissions abatement equipment.
By following this new paradigm, Factory-ECOMATION envisioned model of factory embraces the capability to be: self-conscious, in the sense of being able to perceive, monitor and analyze the whole set of flows characterizing the production environment, and the ensemble of events which change their normal patterns; predictive, in order to optimize its management and control strategy under productivity and eco-friendliness constraints; adaptable at multiple levels to enable the responsiveness to unforeseen occurrences; and sustainable, through the integration of state of the art technologies to increase energy efficiency and reduce emissions.
B.1.1.1 Scientific and technological objectives
Posing its foundation on the concepts previously exposed, Factory-ECOMATION will help European industry to meet the increasing demand for greener, more efficient and higher quality production, fostering the transition to an industry with lower waste generation, energy and raw material consumption.
Framed within this context, Factory-ECOMATION has the following main S&T objectives:
• Developing hardware and software monitoring and data evaluation platform based on unified flows vision.
The first pillar necessary to establish a new paradigm for a near-zero-emissions factory is the implementation of an organic and comprehensive monitoring platform, which could provide in real-time the complete picture of what is happening inside the production environment, taking into account all its characterizing elements, from both the productivity and the environmental impact point of view. Within the scope of innovation of Factory-ECOMATION, this objective will be reached through the development of three elements, two technological and one methodological.
The first of these pillars concerns the design and development of a suitable hardware infrastructure to support all the sensing elements that need to be installed within the production environment. The challenges to be faced in this context are multiples, considering the heterogeneity of these sensors (which, for instance, will have to measure gas flows, temperatures, currents, heat flows, production-related quantities, etc.), the corresponding difference in sampling frequency, the eventual harshness of the installation environment, the need to guarantee effectiveness even under multiple failures, etc.
The second technological element to be considered is complementary to the first one and concern the development of the software middleware required for the coordinated use of the sensor network, and for the low-level filtering and data fusion of the information flows produced by the sensors. In order to effectively enable this software platform, Factory-ECOMATION development will have to satisfy the following requirements and constraints: limited power and resources; scalability, mobility and dynamic network topology; heterogeneity; dynamic network organization; real-world integration; data aggregation.
The third development point is mostly methodological and concerns the design of an adequate data evaluation system, based on the introduction of innovative KPIs who could capture, for the sensed data, high-level information about the environmental performances of the factory (together with those about productivity, which already exist). Within this context, the main challenges will be to include into this KPIs and LCA framework the Unified Flows approach of Factory-ECOMATION, formalizing the potential transformations between the different flows of the production environment, in order to enable the quantitative comparison required for the trade-off analysis.
• Developing specific emissions abatement technologies for the two industrial pilots, covering both external and internal emissions.
The achievement of real-world near-zero-emissions performances in Factory-ECOMATION envisioned factory is subordinated to the development and introduction inside the production environment of emissions abatement technologies, which, for obvious constraints, must be specifically conceived to face the emissions and wastes generation challenges characteristic of the industrial domains chosen for the demonstration phase of Factory-ECOMATION.
Within this context, the solutions developed by Factory-ECOMATION will consider three fundamental aspects: the abatement of external emissions, the abatement of internal emissions, and the total integration with the sensor network platform.
The first point, the reduction of external emissions, is the most evidently pressing constraint to reduce the direct environmental impact of the factory; under this point of view Factory-ECOMATION will act technologically in terms both of absorption or transformation of harmful or polluting compounds, and of filtering and containment.
On the other hand, the element concerning the internal emissions is quantitatively less constraining for the global environmental impact of the factory, but, in Factory-ECOMATION factory vision it is fundamental for the social impact on the workers of the industrial sectors touched directly through the demonstration phase or indirectly thanks to the exploitation and dissemination activities of the project. The objective is therefore to increase the presence of technological equipments within the factory which could act on the working environment, increasing its healthiness thanks to the reduction of noxious elements.
The final point is mostly a challenge in terms of technological development of the project, because, besides the specific design of new technological innovations for emissions abatement, the equipments of Factory-ECOMATION must be conceived as part of the monitoring network established in the factory. This means that all the sensors, and related hardware, necessary to have a clear picture of emissions flows, will be directly integrated in the new machines.
• Developing specific energy recovery technologies for the two industrial pilots.
In a complementary way to the emissions abatement, Factory-ECOMATION will focus its development activities on the design of innovative energy recovery equipments, which will act as the technological intervention on the factory to increase its energy efficiency (the methodological one will be based on the management and control strategies of the production processes).
The same point exposed previously about the total integration with the sensor network is obviously still valid in this context, but for energy recovery Factory-ECOMATION will focus on two aspects: the design integration with the emissions abatement machines, and the multi-oriented availability of the recovered energy.
The first point is aimed at increasing the economic appeal of these technological solutions by proposing a merged approach to the integration within the production environment of energy recovery and emissions abatement equipment: the high-level objective is to push the study of solutions where the two technologies could be at the best merged into one machine, or at least considered as a unified systems where part of the energy recovered is directly used to sustain the processes of emissions reduction.
The second concept that will be developed in Factory-ECOMATION is an extension of the first one towards the total integration of the energy recovery equipment into the global scope of the production environment: within the factory, energy is required in different shapes depending on the final use, this means that the equipments installed to recover part of the energy wasted, must then be able to make it available throughout the factory efficiently and depending on the real-time request. The control of these flows will be left to Factory-ECOMATION management platform, but the hardware itself will be designed taking into account this constraints and consequently proposing adequate solutions.
• Developing a holistic eco-oriented management Decision Support System
Complementary to the technological interventions at the energy recovery and emissions abatement level, Factory-ECOMATION will propose an extensive restructuring of the management decision support systems, in order to include the new eco-oriented performances criteria and consequently take into account and completely adopt the Unified Flows approach introduced by the project.
At this level the main objective will be to provide to the company decision makers an ensemble of tools that could guide the optimization interventions on the factory, be they related to productivity targets, or to the reduction of its environmental impact. In Factory-ECOMATION this will be reached thanks to the comprehensive application on the production environment of the Unified Flows vision: the data gathered through the sensor network, transformed and composed in high added-values KPIs, will be the base for the automatic generation of trade-off analysis of the factory processes, identifying its criticalities and proposing, among the different possible interventions, the best ones depending on the optimization criteria specifically preferred at that moment.
• Developing optimization strategy to support the overall automation platform in reducing generalized factory impact
The last, but not for importance, objective and point of intervention of Factory-ECOMATION, conceived to achieve the near-zero-emissions results, is the upgrade of the entire automation platform of the factory, in order to be able to manage, under the eco-optimized paradigms, both the pre-existing elements of the production lines and the new ones introduced for the sensor network, the energy recovery systems, and emissions abatement equipments.
The main challenge that Factory-ECOMATION will have to face within this scope is the design of a hierarchical multi-level automation and control framework in which energy efficiency and emissions reduction will not be considered just as an additional criterion to optimize, but shall influence the whole spectrum of factory automation systems, from the definition of the tasks up to the choice and execution of the control algorithms.
Under this context, a specific point of intervention of Factory-ECOMATION will be focused on the deeper integration of the simulation tools made available by modern ICT solutions into the design and optimization not only of the production processes but also of their supervision and control algorithms. Through a strict synchronization between the real and virtual production environment, it will be possible to verify the effectiveness of a vast array of optimizing interventions on the factory before having to really integrate them, therefore posing the perfect complement to Factory-ECOMATION Decision Support Systems.

Project Results:
The scientific and technological objectives stated in the paragraph above led, as results, to the following demonstration objectives and scenarios:
• The application of a the optimal control framework to the Brembo furnaces in Poland. Explicit demonstration at plant level have shown the effectiveness of the newly developed technics in improving the performance of the battery.
• The DSS implementation for IKEA test-case, together with the integration with the simulation framework of the project, has allows to validate the methodology adopted and the corresponding technologies as a viable solution in improving the identification of potential maintenance problems and the suggestion of preventive actions.
• The experimental validation of innovative woodworking solutions has been done at SCM testing premises, where a new generation of solutions for aspiration and electricity management for sanding machines has led to unexpected improvements for the overall footprint of such systems.
• The test-rig for low-temperature ORC energy recovery has been deployed in England at Spirax-Sarco facilities, allowing to validate a first generation of devices, pave the way for already envisioned improvements but, most of all, the testing of advanced control algorithms that have shown great potentiality in achieving better ROI.
• The smart monitoring framework of Factory-Ecomation has been deployed at the ITIA-CNR pilot plant, where its ability of simplifying greatly the gathering of shop floor information from heterogeneous sources of data has open the way for future integration of smart systems at shop floor level.
• Finally, the second end-user of the project, IKEA, has continued the integration of the most mature technologies of the project within its plant, leading to results so interesting that the group will extend them in the coming years in other factories throughout Europe.
In the following section, a description of the aforementioned results is presented.
Optimization Platform Validation: Brembo use case
The primary industrial case selected for the validation of the Factory-ECOMATION Optimization Platform is offered by the end-user of the project, Brembo. The platform has been applied to the optimal control of a furnace array of four Inductotherm furnaces located in the Brembo Foundy in Dabrowa Górnicza, Poland.
The platform is composed by three main pillars: a tool for (online) identification; an environment for the modeling of the optimization problem; a set of solvers and APIs for commercial solvers.
Two particular goals have been pinpointed as requirements for the implemented solution: Ability of respect of a hard constraint on maximum power available, with elimination of power peaks and support for real-time energy pricing strategies. A more generic objective pursued and obtained is an improved management of the furnace array during degenerative cases, e.g. a sudden furnace breakdown or a slowdown of the production line. The model has been identified on historical data: a preeminent characteristic of the melting process is that a sequence of phases can be recognized. For each of them, a dynamic model of the furnace has been identified. This approach led to a better modeling of the furnace array, since in each phase a particular phenomenon was relevant: then the most appropriate model has been identified.
The validation of the model, and consequently of the system identification tool, was performed both on a set of historical data, excluded from the identification phase, as well as new data gathered during experimental campaigns. A first scenario was dedicated to the objective of reducing the maximum power request and in cascade decreasing the billing energy price. The model-based optimization computes to which furnace trim a certain amount of power in order to respect the maximum power constraint and in the meanwhile minimizing the deviations of the pouring deadlines and the global energy consumption.
The second scenario is related to the other optimization objective: energy savings. A long tail of the distribution is composed by too protracted melting processes: these melts are characterized by longer and/or additional idle times, after the de-slugging phase. In that condition, the furnace is kept off, full of molten metal. The main reason for this is an absence of enough room in the molding furnace or in the downstream line to accept this metal. During the stopping time, the temperature decreases due to dissipations. Then additional heat must be given to the metal to restore the previous temperature level.
This situation can be overcome with an optimal schedule of the furnace array. The solution implemented with the Factory-ECOMATION Optimal Control Platform, thanks to the internal furnace models, perform a forecast and optimization of the melting process, considering not just the current melt for each furnace but at least two forthcoming melts. The deadlines for them are computed by the application, knowing the production plan. A future additional feature would be an online correction of the plan, knowing the status of downstream production units. An experimental campaign conduced on the shop-floor led to the identification of the energy savings resulting from a better management of the array as a whole.
Decision support system (DSS) and simulation framework
A DSS has been installed and tested successfully in the two Factory Ecomation industrial validators plants; this document is meant to provide a small summary on the DSS implementation and validation phases made during the last year of the Facotry Ecomation project. The developed Decision Support System is a generic instrument meant to assist decision makers and plant managers to optimize their interventions by means of enabling a fast problem diagnosis and a more efficient management of production scheduling, including a convenient allocation of maintenance jobs throughout it.
The DSS is conceptually divided into 3 different modules called Profiler, Assesser and Optimizer that respectively exploit its capabilities of monitoring the shop floor, making predictions on failures and managing the production scheduling.
Being the developed DSS a generic instrument, it is applicable to whatever production system, thus it needs a setting phase before its normal usage. In Figure 9 is represented the part of the DSS software where, in practice, the settings take place; to implement a production line the user has to: Define the number of resources and how they are connected; Aggregate them in terms of workstations whether is necessary; Associate to each resource the KPIs from the list that are available to be monitored on that resource; Provide the nominal values for the selected KPIs; Provide the system with historical data; Provide the system with production data.
The setting usually requires a good knowledge of the shop floor on which the system has to be used; the more the data (i.e. KPIs) are fed into the system the more precise and affordable will be the system behavior.
IKEA industrial implementation
The DSS has been fully exploited in all its modules in the IKEA industrial case; the production line based in Oporto has been fully characterized, leading to the effective mapping of the shop floor in terms of the defined KPIs. In particular the IKEA Industries production line is a very good test bed for the Profiler and Assesser modules because of the high rate of technical problems occurring (so a great quantity of events to be preventively managed) and for the great quantity of production data available from the line supervisor.
Profiler module implementation
The first challenge was to have a deep comprehension of the production line, so to be able to create its model enabling the instantiation of the decisional process defined in the Factory Ecomation framework into the production line.
The instantiation of the decisional process into the IKEA production line lead to the definition of the effective metrics to measure the production performances. These KPIs, that were initially very general and valid for whatever production system characterization, turned into proper indicators for the IKEA production line case. It is clear that, to implement such a process, a deep understanding of the specific case is instrumental.
Assesser module implementation
Once all the metrics have been set, the challenge was to create a big enough database to feed the Assesser module (i.e. the neural networks) with a sufficient quantity of historical data enabling the DSS capabilities of returning back a predictive warning on problems identified in the production line.
After some months of tests has been realized that the minimum quantity of data, sufficient for the optimal exploitation of the DSS capabilities is approximately corresponding to 4 months of production data with a sample time of 5 minutes.
The results coming from the tests showed a great behavior of the Assesser module capabilities leading to reach approximately an accuracy of 98%, thus just the 2% of warnings given by the DSS are not corresponding to the effective problem encountered.
Optimizer module implementation
The DSS Optimizer has been designed and implemented in the IKEA industrial case so to be able to provide two main results:
Static schedule: it is able to provide, with data coming from the Profiler module, an off-line schedule, based respectively on the maximization/minimization of some pre-defined objective functions (e.g. production yield, costs etc.)
Dynamic schedule changes: using information coming from the Assesser module, it is able to modify the schedule on-line, coherently with the objective function, in order to minimize the impact of the problem occurring on the production plant.
In the IKEA case, the static part of the Optimizer module it consists in a sequencing model, designed respecting all the constraints given by the fact of having a transfer line case. Also in this case the deep knowledge of the production line has been instrumental for a proper scheduler design.
Being the IKEA scheduling static scheduler very constrained by the fact of being a transfer line process, the most important part is the dynamic one, where the data coming from the Assesser module (i.e. the warnings on maintenance operations to be done) are aggregated with the static scheduler so to provide a convenient schedule for the maintenance operation.

ORC-unit for low-temperature waste heat flows
The ORC plant aims to recover low-grade, otherwise waste heat into electrical energy, thus improving the plant efficiency and reducing environmental pollution: that means recovering the maximum amount of power, maximizing the exergetic efficiency, not exceeding constraints such as expander maximum speed or torque, complete condensation (since the pump cannot handle twophase fluid) or (almost) complete evaporation.
That said, a conventional fixed-point controller is obviously not suitable for that purpose, because it cannot guarantee neither maximum electric power generation nor constraint satisfaction, especially in perturbed conditions. Actually, a fixed-point controller could be suitable, but only accepting a huge performance degradation in normal operation to correctly operate in presence of “disturbances”: when more heat is available for recovering, a fixed-point controller would try to get back to the design point, and it may recover even less energy than in normal conditions. On the other hand, Model-Predictive Controllers pose the control problem as maximization of a constrained objective, applying the “best” set of input at every control action in order to maximize the objective, so it can adapt its behaviour to the current state of the controlled plant, ensuring always the best working condition.
Approach used to develop the control system for ORC plant based on the Spirax test-rig
In order to build and validate a control system, plant behaviour must be well known. Different ways to obtain a mathematical model of the plant exist, however they could be distinguished in two main approaches: first-principle modelling (i.e. deriving differential equations from principles like conservation of mass and energy), and black-box modelling (i.e. building an abstract and conventional mathematical model completely detached from physics). Although the latter is generally faster to build, easier to control and better for online identification, it misses knowledge of plant internal behaviour and could mask faults, whereas the former is more challenging to build, simulate and identify, but at the same time allows a deep understanding of plant behaviour for both humans, who can design better controllers, and MPC, which can diagnose sensors/actuators/devices faults by an accurate model.
The ORC system is composed, as other Rankine-cycle based plants, by an evaporator (optionally ), which heats up the working fluid (in this case R245fa) until it is totally evaporated, followed by an expander (turbine, screw expander, scroll expander, ecc.) which effectively converts fluid enthalpy into mechanical energy (converted then into electricity by a mechanically coupled generator). The fluid leaving the expander enters a condenser, which causes the big pressure drop needed by the expander, and then enters a decoupling receiver that ensures the fluid entering the following pump is liquid. The pump then increases fluid pressure before entering the evaporator, so closing the cycle.
In the Spirax-Sarco test-rig there are other devices primarily intended for warming up and cooling down the plant, such the bypass pump and the bypass valve: these let the fluid circulate until heats up to vapour, because the expander is not able to tolerate low-quality fluid (i.e. with a “low” percentage – under 90% - of vapour; not related to quality as in “product quality”).
A first-principles model of the ORC plant has been developed and identified with good accuracy. A control system has been developed and simulated and has shown a suitable behaviour; however, it does not take in account startup and shutdown, due to some inaccuracies and very strong nonlinearities in fully liquid conditions, nor enforces constraints like maximum expander speed. Furthermore, control speed response could be improved by tuning a Model Predictive Controller, which can push the plant towards its limits satisfying constraints at the same time.
Following steps will improve model accuracy and develop a Nonlinear Model Predictive Controller, that will be extensively simulated and then deployed to the system. It is advisable to build also a first-principles-based data reconciliation procedure to cope with unknown sensor offsets and drifts.

Smart-monitoring middleware to the remanufacturing pilot plant
The main element of innovation of the Smart-monitoring middleware is in fact the introduction of a global factory virtual counterpart which has to represent the object through which all the needed information can be provided in the most fast, easy and efficient way to high level applications. This factory virtual counterpart presents the following basic features:
• It has dynamic data storage capability exploiting an intelligent approach to organize the information required and nevertheless to automatically deal with the factory database management.
• It is synchronized to the real factory in order to provide always the most updated information related to every aspects of the production environment.
• It provides an effective standard way to access to the required information by implementing a service oriented approach to allow the information exchange between heterogeneous objects.
In such a way the sensing and monitoring platform is the way to easy access, through standard interfaces based on web services, to all the factory data. Furthermore, this could enable the realization of a next generation of high level applications that could expose to the user a complete collection of information, allowing for example the possibility to create new and more complex KPIs by aggregating different types of data coming from the different areas of the factory.

Revamping of Dust Extraction and Filter systems
The focus is on dust extraction due to electrical consumption, noise and poor chip capture efficiency.
During the time frame of the Factory Ecomation project, the electrical consumption of several factories have been mapped to understand and the share of consumption for different production processes and support systems. It has become obvious that the dust extraction system is an area to prioritize. Typically we have 20 to 40% of the electrical consumption for this support system, which is more than the consumption of the production lines they are connected to. Dust extraction systems are also one of the main sources for noise inside and outside the factory. The high noise level is seen as a main contributor to make the factory environment a less attractive place to work. The deficiency in the chip capture efficiency, meaning the inability of the system to collect all dust and chips that are generated by production process, creates a dusty atmosphere, which may have a negative health aspect, and it generates a lot of line stop time for manual cleaning. In a study that was made in the Paços de Ferreira factory in March 2015 where a batch of 6035 panels where produced during approximately 4 hours. All dust that was left in the machines and around the machines was collected and compared to the calculated theoretical amount of dust that was generated. In total we could conclude that 98.6% was collected. That doesn’t sound too bad! In the production batch 3656 kg of chips and dust generated. Left in the machines and production hall was 50 kg of dust. In three-shift operation we normally need to stop the line for 1.5 hours/day for cleaning.
The use of cutting tools that are designed to use the kinetic energy of the chips to move the chips into the extraction hood. The chip emission angle shall be well defined to enable efficient design of the hood. In some cases the chips are very efficiently extracted through a hollow core of the tool.
Extraction hoods that are designed using computerized flow and particle simulation to optimize the performance. The design of the hood is very closely connected to the tool design (mainly chip emission angles and air flow). The combination of hoods and tools have the largest influence on the chip capture efficiency. They also have a large influence on the energy consumption since the required air speed and under pressure level is defined here. By designing hoods to equalizing the demanded pressure and speed throughout the line, the system set-point can be reduced.
Connection of the hood to the main ducting with pipes and tubes with the lowest possible pressure drop. In this part of the system often flexible tubes need to be used and they often cause high pressure drops. Various samples have been compared with huge difference demonstrated between them. On top of the machine all hood connections come together in a manifold collector structure. This structure must be well designed with calculated area increase and appropriate bends.
Duct design with correct dimensioning and design. The duct work must be designed perfectly for the air volume it shall transport, and in a way to avoid unnecessary pressure drop. The duct dimensioning is very critical as the air flow should be kept to a minimum but ensuring transportation of chips/dust and yet compromise dimensions to have lowest possible pressure drop in the duct system. Too low speed and flow will cause material build up in the ducting and create risk of clogging and further conflict to the ATEX (explosion) regulative.
The filters must be designed with efficient cleaning systems, lowest possible pressure drop and allow for a large range of filter area loading. Filters should be designed to handle peak loads in the system in order to avoid clogging and by that heavy down time of the extraction system.
The fans must be designed with impellers that are specifically optimized for the working conditions, by that the fan will be efficient within their specific working range. The fans must be installed with inverters to control the speed.
The use of dampers, to shut of parts of the system not in use. To save energy by closing dampers, the fan must be automatically adjusted. Safety must be respected as the ducting can implode if too many dampers are shut down too fast.
Signal exchange with the machine to only run the system when it is needed. Stand-by signal from the machine enables to turn off the fan when there is no production. Signal communication between machine and extraction system will allow dampers to react (open/close), and the speed regulated fans will adjust extraction to a minimum requirement.
An on-demand control system will monitor pressure level and air speeds in critical positions of the system, and minimizes the air volume for all scenarios based on several principles and priorities.

Potential Impact:
For each of the identified project output, a dedicated business model and a tentative business plan has been defined, this highlighting the exploitation path identified for each result.
In particular, the Factory-Ecomation Project has reached as project output, 7 exploitable results that are expected to enter the market in average, in two years after project end. The results are built on top of a hierarchical reference framework, composed of many different composing and complementary technological elements, at equipment, software and methodological level, and in particular:
1. Distributed automation platform
2. Predictive control engineering framework
3. Smart monitoring middleware
4. Eco-efficient woodworking solutions
5. Simulation and decision support framework
6. Model-predictive emissions control
7. Low-temperature ORC units
In the following section a brief description of the impact of each result is presented:
1.Distributed automation platform
Industrial automation is changing, driven by the increasing digitalization and cost pressure. Intelligent production will be based on more and more distributed intelligent devices, on more and more accurate data provided to managing systems and on an engineering able to master such complex systems. Flexible control on factory floor and seamless integration to IT-systems will help to make production more intelligent and more efficient (very much in consumption of energy and better handling of waste). As a summary we answer the need for:
a) Efficient solution for distributed control systems (hardware, software, engineering)
b) Innovation but with ability to re-use of existing solutions and the ability to address installed base
c) Cut costs (time and equipment) for testing and commissioning, both for new equipment and as well for reconfiguration
The products are based on the IEC 61499 standard for distributed systems, by that the basis for easy distribution of control logic is in the DNA of our entire offer. The engineering tool we have developed integrates seamlessly several tasks of automation but especially enables and easy integration to IT-systems and offers the re-use of existing solutions. As a summary we offer:
a) IEC 61499 based control software to answer the increasing need for more intelligent devices distributed in equipment and plants. The advantage offered by nxtControl: easy distributable control logic and an engineering tool mastering complexity of distributed systems
b) Extension of product features by integration of C++ based algorithms encapsulated in function blocks as used in IEC61499 engineering environment. Optimization and other advanced algorithms can be integrated easily and existing C++ solutions can be re-used without limitations.
c) Virtual simulation and commissioning integrated to our solution will help to reduce costs in many ways. Time for commissioning will be reduced massively as part of commissioning can be done from your office desk, commissioning can be paralleled, simulation helps to get results of optimization or reconfiguration means in advance of physical implementation, scenarios can be evaluated in an virtual environment but connected to physical equipment, etc.
d) The increasing number of intelligent devices in a distributed network will ask for more cost efficient controllers and electronics, which are small and can be implemented in different form factors. The project has helped us to develop a control hardware perfectly fitting this needs. A hardware with reduced CPU power, small memory but able to execute even advanced algorithms and control tasks. The hardware is offered as a standard device or as a system on board solution.
The intorduction of this solution faces several different kind of competitors as we are offering an integrated solution. One of them are the control software manufacturers relying on the IEC61131 standard. They represent big part of the existing market. This are companies like 3S (Codesys), Isagraf, KW-Software, Infoteam, etc. With their technology based on IEC61131 they are not having an appropriate solution for the needs of flexible manufacturing with distributed intelligence and easy integration to IT-systems. Other competitors are coming from the HMI/SCADA side moving down to the field control level. Most of them are very spezialised in niche markets. The bigger ones are already part of big international automation vendors. The third an most difficult part are the global players in the automation world like Siemens, ABB, Schneider Electric, Mitsubishi, Honeywell, etc. They are difficult because they are our targeted customers as well. They are facing the problem that their solutions might be efficient and comprehensive but always proprietary and by that not flexible. More and more end-customers try to reduce the dependency, which is created by this proprietarity.

2.Predictive control engineering framework
The markets for the new equipment of F-E will be the same of the usual one for CTP, that is Automobiles and Vehicles, Chemical Industry, Petrochemical Industry, Construction Materials, Consumer Goods, Electronics and Electrical Industry, Energy & Resources, Food & Beverage, Health Care, Metals and Mining, Oil & Gas, Packaging & Printing, Wood, Paper and Fibre Boards, Cement Industry. In order to better satisfy the customers, they are grouped in distinct segments with common exhaust gas problems. Such groups are e.g. Metal and Mining industry or Wood, Paper and fibre board industries as defined for F-E, but also other ones like chemical industry, electronics and electrical industry.
The main aspects solved by the new optimized emissions control equipment is the sensible reduction of energy consumption, that can be used either to reduce the costs of usage for the end-users or, at equal consumption, to obtain much higher filtering capabilities. The optimal emissions control system is capable to obtain this increased performance because it adopts a new design of the combustion chambers, conceived by CTP as a result of the project, that reduces the losses throughout the catalytic process. Morevoer it implements a model predictive control of the hybrid non-linear processes that is capable of reduing the inertia of the system and, consequently, reducing the peaks during the cycle inversion.As a consequence, the equipment as whole is capable of maintaining the same level of emissions filtering but with a reduced average temperature of the combustion chamber, leading to an obvious decrease in fuel use to heat the chamber.
The main market competitors are already existing produces of emissions filtering equipment, advanced system integrators capable of customizing existing systems with new type of control systems, developer of optimal control solutions that could be installed on other similar equipment.
Since the market initially involved with the solutions implementing the new technologies will be around the 10% of the portfolio of solutions of CTP, estimated market size is of around 10 M€, with a 5% growth rate per year.

3.Smart monitoring middleware
In the industrial environment there is the need to gather data to picture processes. In addition the data is needed to build KPI's. In practice there are lots of heterogeneous data sources because of the existence of different systems and underlying protocols installed within a site. To be able to build KPI's from the different heterogeneous data sources it is necessary to provide a wide range of protocol adapters to access them. The number of different protocol adapters isn't limited.
The middleware offers an internal interface for new protocol adapters, which may become necessary during the technical improvement and further development of a customer’s site. Over time new protocol adapters can be included into the existing system easily to meet the demands. Furthermore there is the need of a consistent system which manages the incoming data, provides the capability of long-term storage and calculation of derived indicators. Also the system needs to provide a standardized interface so high-level application can access data without the need of complex data transformations. The way of providing data isn't fixed and can be changed and extended.
The product (energy management software) will hide the complexity of dealing with heterogeneous data sources. The user can use existing protocol adapters, encourage or order the development of new protocol adapters or can encourage changes of the existing ones. So the user is prepared for further technical improvements and developments within his site. Because of the standardized extensible interface of the middleware the user can introduce new high-level applications later if needed. There is no risk of incompatibility between existing products and new ones that need to be integrated.

At Germany there is a list of eligible energy management software from the Federal Office of Economics and Export Control ( This list is shows a good overview of competitors. The competitors are different related to special topics of the scope of energy management as well as related to the company's size. Therefore it is difficult to concentrate on specific competitors.
A former version of the product is installed at nearly 250 sites all over the world. The new version is currently installed at 5 sites at Germany. Because of law and the general interest to invest in energy efficiency and cost reduction the market size is big. All energy intensive industrial sectors and non-productive industries which have the requirement to build KPI's from different heterogeneous data sources for optimization are potential customers. Especially energy intensive industries with a high ratio of energy related costs to manufacturing costs can profit from a holistic factory view supported by KPI's. At present the market also includes other participants that were not in the focus before e.g. banks, public authorities and hospitals.

4.Eco-efficient woodworking solutions
In the last decade SCM implemented more and more integrated cells and production lines including different technologies such as boring, edgebanding, drilling, sanding units ecc. Several customers and several national rules are asking to reduce the energetic impact of production systems. This need is related both to energetic costs an to environmental constrains.
In the previously described context two main actions, related to Factory-ECOMATION results, are envisaged: ECO-ORIENTED RETROFITTING related to “Modelling and acquisition of a formalized know-how on energy-related issues in real production context of wood manufacturing to allow a wide set of smart retrofit actions on existing production systems for panels machining” A wide impact on market of existing panels production systems thanks to an expected saving of around 20% in Electrical energy and compressed air savings NEW ECO-OPTIMIZED MACHINES related to “evelopment of modular components and automation functions for breakthrough new generation of woodworking sanding machines with very low energy/environment impact over their life cycle”. Outdo competitors in the supply of sustainable solution for sanding machines thanks to an expected saving from 10% to 65% (depending to product mix) of compressed air savings. The concepts and specific subsets (control algorithm and/or sensors) of the solution implemented for sanding machines will be extended to other specific machines depending on the cycle characteristics.

Woodworking machinery market is very competitive and SCM main competitors are companies from Italy and Germany and USA. Specific actions related to energy saving and emission reductions have been addressed by main competitors like Homag Gruppe and Biesse Group. In this context the specific solutions that SCM will to implement, especially in retrofitting solutions for integrated lines, will be allowed by a strongly modular approach introduced form year 2000 in SCM design team. Such approach today is under continuous improvement (modularity and interfaces) and can be considered one of the key to increase the competitiveness of SCM proposal (ECO-ORIENTED RETROFITTING) on the market. Moreover SCM have a very wide set of customers in all market segment and this can support the extension in the next years of NEW ECO-OPTIMIZED MACHINES action starting from mid-high level sanding machines typically integrated in complex production lines.
The project results will focus on a wide market size that is related to: 100 % FURNITURE INDUSTRIES for ECO-ORIENTED RETROFITTING OF EXISTING PRODUCTION SYSTEMS FOR PANEL MACHINING. The typical use/need that the solution will cover will be: (i)Production lines with high capacity and automation for panels machining, (ii)Energy consumptions for single machined piece as KEY PARAMETER. The project results will focus on a wide market size that is related to:
70 % PROFESSIONAL WORKSHOPS and 45 % FURNITURE INDUSTRIES for ECO-ORIENTED ADVANCED SOLUTIONS FOR A NEW GENERATION OF SANDING MACHINES. The typical use/need that the solution will cover will be: Occasional use of the machine with frequent changing of work-piece dimensions or Intensive use of the machine, medium/small production batches up to batch 1, frequent changes of work-piece dimensions.

5.Simulation and decision support framework
In complex manufacturing environments hosting various technologies and manufacturing stages of the product lifecycle as well as different production systems, the decision making process is severely complex. The decision making process involves multiple divisions and production system logical layers. The only two reasons that can be identified for stopping a production line are due to regularly scheduled maintenance or equipment failure. The decisional process associated with maintenance activity on manufacturing lines becomes thus a critical activity, if considering the repercussions that stopping the production line can have on the whole plant. Performing timely and necessary maintenance is indeed critical to preventing failures that may result in costly production interruptions, but relying on a fixed schedule may result in higher than necessary costs for both parts and labor. In many large-scale plant-based industries, the costs related to the maintenance of such level of availability can thus account as much as 40% of the operational budget. Moreover, up to one-half of these maintenance costs can be considered wasted by the application of ineffective maintenance management methods.
Despite the state of the art progresses in the application of sensors and tools able to track machine behavior and performances, organizations are still lacking in a structured and reliable approach for collecting and analyzing equipment performance, executed maintenance tasks, failure history or any of the other data that could, and should be used to plan and schedule tasks that would prevent premature failures, extend the useful life of critical plant assets, and reduce their life cycle cost. On the other side, when data are extensively tracked, most of the times obtained information are stored in large databases with very limited possibility to effectively use them for analysis and correction purpose. In this context, the ability of collecting and analyzing production lines’ problems becomes instrumental in order to be able to implement effective programs supporting the early identification and prevention of machine problems on the lines. The importance of being able to perform data analysis and forecasting decisions has been also confirmed by a recent Accenture survey of 600 business executives that highlighted how the use of forward-looking data analysis tools has tripled since 2009, and by a Transparency Market Research report anticipating the market for predictive analytics software to reach $6.5 billion by 2019.
Considering that the alternatives decisions amongst which a decision must be made can range from a few to a few thousand, decision-maker needs a supporting system able to narrow the possibilities down to a reasonable number. Decision support, such as a selective information retrieval system can help with this task. In order to be effective, the analysis and decision making process must be structured and implemented considering at least the following steps:
1. Data acquisition. Organization of predictive data gathering with the support of lines’ sensors and machine operators. The data gathering activity has to be defined, by operator point of view, as an activity that minimizes the necessary time to be performed.
2. Information processing. Once the predictive data has been acquired it must be transferred to PC or a workstation to be processed. The data transfer must be carried out automatically each time a measurement route has been completed in order to avoid any lack in data consistency. Supplementary registers can be acquired or a more advanced analysis can be applied if deviations in the behavior are discovered.
3. Interpretation and drawing up of diagnosis. This is the key activity of the decisional process and should be carried out by qualified personnel whose principal objective is to avoid the accumulation of unanalyzed registers and predictive samples. Using the acquired predictive data, an analysis of the information is carried out for which the expert uses as an aid graphic representation of the data. In order to be able to drive appropriate decisions on the state of the plant/line, a software able to manage the structured data and to propose possible preventive intervention has to be of support. The technician using these tools, in addition to his experience, is thus able to establish the state of the equipment and if this is not satisfactory issue a diagnosis of the failure. The technician has at his disposal an expert system, which using some preliminary data, issues a diagnosis of the state of the equipment.
In order to solve the aforementioned issues, an Integrated DSS-Simulation tool providing the following benefits is the core product offer, this providing :
• Early Identification of problems arising on the production line;
• Support of preventive maintenance in manufacturing plants, by suggesting required maintenance
• Rescheduling of production plan by considering the foreseen problems occurring on the line
• What-if analysis of updated scheduling plan through line simulation
• Monitoring and analysis of line’s/resources related indicators
• What-if analysis of line’s KPIs with different preventive maintenance plansCompetition
The tool has been developed in order to be general enough to be applied to any production line. Main customers are mid-large manufacturing industries where complex manufacturing systems are present. The only requirement that could restrict the application of the tool is the necessity of retrieving a considerable amount of data from the line (i.e. proper amount of sensors on the line). In addition to the availability of the information we should also have an historical data base containing those information, the historical data is needed to train the system.

The cost for evolving the DSS prototype to an industrial application is estimated as ranging from 40 k€ to 80 k€ (6 MM) for development, this considering the GUI redesign, the AI algorithms refining, implementation and final testing.
The cost for evolving the Simulation Framework prototype to an industrial application is estimated as ranging from 50 k€ to 80 k€ (6-8 MM).
In order to cover these costs and be able to more sustainably reach the market, further development costs are meant to be covered by participating in a following project intended to enable the final steps required for the reaching of the market.
In the following scheme, the expected development plan is presented.

6.Model-predictive emissions control
With the context of the Customer segments identified for Melt.Brain (iron casting foundries adopting electricity-based induction furnaces), due to very specific technical requirements that the product has, the following user profiles have been identified as interested to the solution because of its potentiality to solve their needs and reduce their problems:1.The production planner, that is, the person (or people) with the task of deploying a high-level production plan of molten metal for the array of furnaces in order to satisfy the requirements of the production line that is fed by such furnaces. 2.The shop floor supervisor, the technical operator that, acting at level of the array of furnaces, transform the production plan for the full array into single sequences of production batches for the furnaces.3. The furnace operators, the people that concretely control the furnace, throughout the phases of its melting procedures, to assure that it works correctly, at its maximum possibilities and satisfying the production targets imposed by the shop floor supervisor.
Correspondingly, for these three categories the following main needs/opportunities have been identifies: 1.An improved awareness of how choices at production planning can influence the productivity of the array but, most importantly, the opportunity of forecasting the effects of such choices in a quantitative way. 2.An improved and more automatized way of taking into account the real behaviour of the array, in particular its non-nominal events, when dealing with the transformation of production plan into concrete batches for the single furnaces. Moreover, the opportunity of improving the efficiency of the furnace by optimizing their management. 3.A simplified way of taking into account the effects of choices on the other furnaces (by other operators) on his own.
Melt.Brain is the solution to solve the abovementioned needs while answering the corresponding opportunities because:•Thanks to its model predictive technology and the integration of non-linear simulation engine synchronized with the real-data of the plant, it allows to create easily what if scenario on different production plans;• It computes optimal distributions of production batches to the furnaces in order to minimize waste of energy while satisfying all production and electricity constraints.
• It is completely integrated with the operator HMI, providing real-time indications about how steering the operator behaviour in order to follow the abovementioned optimal plan.
The main market competitors are the followers:
•Producers of general purpose plant management systems, such as the Siemens product adopted by Brembo, that does not solve specific furnace problems of the array of furnaces but are fully customizable and programmable and, as such, could be used by consultancy firms to re-produce the same results of Melt.Brain (something that would require the corresponding competences and development effort of ITIA-CNR).
•Control software of the induction furnaces, such as produced by Inductotherm or FOMET, that could be improved to adopt predictive control algorithms, but without having the possibility of looking at the problem from the production planning perspective that Melt.Brain can offer.
•Software conceived to do peak shedding of the power absorption, already present on the market and in partial overlapping with the value proposition of Melt.Brain
The model of pricing for Melt.Brain is linked to the number of furnaces to be managed, generally from 2 to 8. Therefore the market size has been considered based on the number of installed induction furnaces at European level, which is around 20.000.

7.Low-temperature ORC units
The proposed solution is a low temperature Organic Rankine Cycle (ORC) system that will take low grade energy below 100°C and generate electricity.The primary target will be any customer that has processes that require cooling, where the cooling water exits the process and is routed to a cooling tower or chiller to be cooled and then routed back to the processes to provide further cooling.Our application target will be where the water flowing to the cooling tower is in the temperature range of 70°c to 90°C, and the thermal cooling load is between 1.3 MW and 2 MW.After a full investigation and testing of the ORC system, a slightly modified system will be developed and sold, known as a Trilateral Flash Cycle (TFC) system, which will deliver much higher efficiencies in terms of kW(electrical) output per kW(thermal) input, and therefore will improve its commercial viability.A number of ORC system suppliers are already in the market, but there are none who can address applications where the incoming evaporator temperature is as low as 70°C, and only a few that can successfully address 90°C and still be commercially viable.Therefore the market for energy recovery and electricity generation solutions for streams below 90°C is virtually unexploited.However we must remember that we are not only competing with other ORC offers in this particular application, but with any other energy saving and investment opportunities the customer may have.
The attractiveness of a market is largely dependent upon the cost to the customer of electricity. A secondary factor is the availability of energy of investment grants or ongoing incentives for saving energy and/or reducing carbon footprint.
An initial market assessment has shown us that in the power range of 1.3MW to 2MW, in the markets where Spirax Sarco has a direct sales presence, there is a total population of approximately 500k cooling towers. From this we have extrapolated a Potential Available Market PAM) of 18k cooling towers that we can target for this energy recovery technology.

Informations connexes


Paolo Pedrazzoli, (Senior Researcher)
Tél.: +41 058 666 6638
Numéro d'enregistrement: 182470 / Dernière mise à jour le: 2016-05-12
Source d'information: SESAM