Final Report Summary - SUSTAINHUB (Sustainability Data Exchange Hub)
Executive Summary:
A number of new challenges for companies in the EU emerge in a globalised economic cycle.
Manufacturers and their suppliers from the electronics and automotive industries are required to comply with restrictions for selling and exporting products in the EU.
Directives like the RoHS and ELV interdict the use of certain substances. In addition, the REACH legislation requires companies to conform to a whole range of requirements related to production, import and use of materials.
Adding to this is the need to comply to the increasing wish for sustainable products / production, be it by own choice or society pressure. The reality in industrial production and trade is way behind this. Main causes for this are the lack of trustful information about sustainable indicators between suppliers and customers (caused by a missing or incomplete data exchange or missing standardization) and the insufficient knowledge, especially in smaller companies, about all sustainable issues, be it regulations or laws or possible threats emerging from wrong conduct of these laws and regulations.
SustainHub therefore combined organisational and institutional measures with technical solutions.
First of all SustainHub addressed an important factor which is a bare necessity to implement a lasting new service, a business model, which is as well stable as scalable. This business model points out a way to implement the SustainHub data exchange as a self-financed service. For this service SustainHub defined a set of valid sustainability indicators (and methods to calculate and aggregate them) and created the necessary IT tools, namely advanced communication facilities for supply chain information interchange, a data model capable of managing all aspects of product and related sustainability data, tools to expand the available information basis by data mining principles and tools to improve the data quality by using algorithms for automatically checking the plausibility of data.
Project Context and Objectives:
The Challenge
European companies are increasingly under pressure to bring sustainable products to market. Especially high technology industries such as electronics and automotive are driven not only by environmental legislation but by consumer demand to produce sustainable products. Companies trying to address this situation face an information problem: exchanging compliance and sustainability data on corporate, component and material levels across the entire supply chains. The prevailing situation in the industry is that participants take a reactive position regarding compliance and sustainability data exchange.
The motivation for OEMs is to avoid costly product recalls and public relations crises. Therefore they pass the data exchange requirements on to their supply chain, and their market power forces suppliers to comply. Suppliers in the middle / lower supply chain usually have smaller company sizes and they are under pressure to meet the data requirements, in order to assure their market access. Their role is especially resource-intensive, as they have to gather data from several other tiers and process them into required, often custom reporting formats.
In consequence those companies, which often lack dedicated resources, suitable processes and IT systems for sustainability data management, are struggling to fulfil the customers’ demands.
As a result are the current processes concerned with sustainability data processing inefficient and largely manual and necessary data is either incomplete, failure prone as well as stored in a multiplicity of formats and sources. The integration into operational internal processes is insufficient, exchange processes are slow and complicated.
Project Objectives
In the long run SustainHub will provide an open and self-augmenting collaborative marketplace for various sustainability data processing topics. Besides the overall concept for the SustainHub marketplace, the project planned to develop first fundamental services to be offered on the platform. Those services are dedicated either to increase data quality, risk and information transparency or to lower manual effort and therefore overall costs associated with the processing of sustainability data.
Regarding data exchange the objective was the development of an easy exchange solution of eco-/sustainability data along and across supply chains. For this purpose the SustainHub project had to define organizational measures like procedures to gather and aggregate sustainability data and furthermore had to establish an IT-supported collection and transformation (e.g. from one standard to another) process. The link to present databases should furthermore enable the reuse of existent data. To further support data collection within the companies, a support function had to be setup to guide inexperienced users through the collection process.
To address data actuality, correctness and quality issues, the projects objective was to provide first services which enable potential users to receive early information on necessary data as well as to check generated data.
For this, the SustainHub project planned to deliver a tool which enables the user to identify upcoming sustainability requirements in a very early stage. This shall allow companies to integrate new requirements in an early phase of the product development.
Above that, plausibility procedures are to be obtainable, which evaluate the sustainability data quality. These plausibility checks are to be supported both by checks from a professional (human) perspective as well as by automated checks based on data mining techniques. Those techniques will additionally be used to estimate data gaps and to establish an automated risk assessment on in-house sustainability data.
A collaborative testing tool had to be developed, which enables companies to share sustainability tests (e.g. XRF-tests for RoHS-Screenings).
Finally, the development work is encompassed and guided by a Business Model which shall ensure the future market access.
Project Results:
The scientific and technological results of SustainHub are:
S/T I: Collaborative Sustainability Network along the entire supply chain in order to enable the easy exchange of eco-/sustainability information along the supply chain. This means, the development of a business model for a self-augmenting collaborative network for sustainability data.
A comprehensive survey of the industrial requirements was conducted. The gathered data / requirements have been analysed in absolute numbers as well as dependent on the priority in each customer segment. Findings of the quantitative and qualitative surveys were documented in D1.1 and first conclusions for the implementation of the SustainHub platform were drawn. This included the definition of preferred functionalities and features. The main results for this part were
- Customer expectations regarding all steps of sustainability data processing from collection through data preparation and analysis to data transfer. Each result has been specified according to tier position and company size.
- Importance of different sustainability topics and their relevance for data exchange in the supply chain and the efforts linked to the data processing identified.
- Information regarding customer segmentation, incentive and pricing strategy available
The Identified relevant sustainability issues were analyzed for their complexity. Issues like type (qualitative or quantitative) and format of data, completeness, transparency and actuality of data as well as external verification of data were considered. In order to test and validate results, experts with profound experience in supply chain data exchange in the automotive and electronics industries were consulted to test and assess the organizational data requirements. Finally the input of experts in the field of sustainability management was taken into account to select the most appropriate type, diversity and amount of data for the indicators. Lastly the requirements were evaluated and defined from an organizational and technical point of view. A case example of a manual, which describes the presentation of quantitative and qualitative data, was established. For this purpose gender and energy related data were selected. The main results for this part were
- Requirements arising from sustainability data are defined (complexity in terms of diversity, type, amount).
- Necessary sustainability raw data is defined as a basis for the development of the Data Model.
- Manual for the identification of environmental and Sustainable raw data.
- 36 sustainability indicators are selected (out of 115 aspects) for further processing in the SustainHub project.
Based on the output of different workshop sessions and research done by the project partners a common vision on the business model has been defined. Different alternative options and solutions have been formulated and prioritized. This makes the business model flexible towards future changes. Due to interdependencies with the technical requirements, several specifications of the business model have been adapted in several feedback loops. The main results for this part were
- A Business framework and a complete roadmap for the successful introduction of SustainHub into the market is concluded.
- The Customer segmentation and key players in the field of concern are defined.
- Overall value proposition of the marketplace and the specific services by addressing each customer segment is specified.
- Necessary resources are calculated.
- A financial model (pricing and costs) is established.
- Marketing strategies are deducted.
- An incentive model to guarantee high participation (with special consideration of SMEs) is specified.
The requirements for the SustainHub business model and for the technical concepts have been consolidated from the requirements and documented in D1.4. The technical requirements have been transformed into user stories and prioritized, in order to scope the prototype implementation. From this a detailed list of technical use cases with implementation specifications has been elicited from the user stories. This specification sheet has been designed in such a way that it could be utilized as guidance for the development of a prototype. The developed business model has been evaluated with consortium partners and interested third parties. The main results for this part were
- Consolidated requirements with impact on the technical solution.
- Prioritized user stories for technical solution.
- A specification of technical uses cases for the implementation.
- A roadmap for the technical development.
- Evaluated assumptions (by potential industrial customers ) of the business model and the technical concept.
S/T II: Scouting tool for eco- and sustainability requirements for their rapid integration in the design process of products and the avoidance of expensive redesign cycles. To reach this objective, reliable tools for the early detection and recognition of changing or new eco-/sustainability requirements are necessary.
A list of product characteristics with significant impact on the eco-efficiency of products has been gathered from various guidelines and articles. This list has been structured and a general target and lever system for eco-efficiency from a product perspective has been developed. This can be easily adopted in product development and for example in the customer requirements survey. Where possible the list was supplemented by a set of international standards for each characteristic. Based on the presented system a methodology was developed which describes – on an aggregated level – how certain upcoming requirements related to the defined product characteristics could be detected precociously on the internet. Defining the according sentiments linked the model to the technical implementation of the Eco-scouting tool. The main results for this part were
- A list of product characteristics with impact on the eco-efficiency of products. This list gives a well-defined orientation on what kind of characteristics need to be taken into account in product development.
- A procedure how to precociously detect requirements regarding eco-efficiency of products on the internet.
- A list of sentiments.
- An eco-scouting tool – search and analysis engine for material related product characteristics (e.g. Hazardous Substances).
S/T III: Collaborative Testing Network in order to save resources for testing components and materials (e.g. RoHS issues) a common use and exchange of the results is planned. This network will reduce the number of needed tests and efforts for every single member while the data quality is improved. The benefits of using the testing network are they will strengthen the dissemination of SustainHub.
The base for the Collaborative Testing Network (CTN) is the need for reliable storage and retrieval of test data. This includes the mapping of test results to components or materials. A relational database was concluded by comprising of all necessary data entities (demographic data plus BoM and all risk influencing parameters). Especially the intelligent testing procedures, material classifications, material properties, categorization of legal requirements and chemical elements have been defined and prepared according to the database design. Several thousand real data sets have been imported repeatedly and tools for the import from other databases have been established. Based on the developed use cases, rules for the risk assessment are described and have partly been programmed. A concept for the organization of collaborative testing has been created. This allows the participating companies in the network to decrease their operational risks of non-compliance and also reduce their efforts and expenditure on testing. The developed data base and user interface has the potential to help companies to ensure the fulfilment of their due diligence as required by the regulation (EC) No 765/2008 (market surveillance). The main results for this part were
- A concept for intelligent testing.
- A model of risk influencing parameters.
- A relational database is operational, import tools are programmed, the database is filled with necessary data (regulations, substances, examples for BoM).
- A concept for risk calculation is defined and can be demonstrated.
- A concept for the organization of collaborative testing is defined.
One aspect of the business model mentioned above was the future integration of further service partners into the SustainHub. In a first step different kinds of possible future service providers were defined (e.g. content providers, service providers like consultants). This resulted in a general qualification of possible future services along the steps of data processing. Furthermore the integration process was described in terms of quality management using a QFD driven approach for requirements analysis and revision processes. Finally, an according business and incentive model for future service providers was added in order to lay down how they could be attracted to the SustainHub platform. The main results for this part were
- A method is described how additional services can be integrated into the platform.
- The qualification of services is defined.
- Quality management procedures are set up for the integration process.
- A business and incentive model for additional service providers is defined.
S/T IV: Methods for validation and estimation of missing eco-/sustainability data, within this objective the development of methods for validation of existing and the estimation of missing data will be realized. Using these methods will improve the quality of the data and enhance the acceptance of the data network.
The areas in the database, in which data could be missing / be false and what kind of data types could be affected in what way were defined. A list of queries for the data estimation / validation was created which shows the different problems and goals regarding missing / false data. Based on real data sets different validation and estimation strategies have been developed. In summary, the data mining algorithms could be successfully applied to achieve the use cases for the SustainHub platform. Supervised classification techniques outperform the association mining techniques and also supervised classification techniques could be successfully used to achieve the use cases for missing data and plausibility checks. Though there were challenges due to limited data access, the results obtained for the substances that are available in the prototypical data look promising. With the availability of more data, we expect the system could perform better and the reliability on the results from data mining can significantly improve. The main results for this part were
- A list of use cases / queries for the data mining.
- Different mining strategies and procedures are applied and evaluated.
- Several data patterns are identified and documented.
S/T V: Identification and definition of environmental and sustainability data. A concept for the quick and easy integration of changing or new requirements concerning eco-/sustainability aspects in the supply chain will be needed. Based on case examples, a code of conduct will be determined and defined.
The sustainability aspects within the supply chain were evaluated through a comprehensive literature research, the evaluation of numerous sustainability reports and by interviews with industrial partners. A deductive quantitative and qualitative content analysis was applied on the interviews. The deductive approach was used as a ranking criterion for identification of the most relevant sustainability aspects. Due to the combination of both approaches, more reliable and significant results were achieved. Finally, a quantitative online survey was designed in order to serve as an additional filter for the identified sustainability aspects. The survey also helped to better understand the relevance of sustainable supply chains for various stakeholders and was available in English, German and French. The aim was to gather approximately 100 answers from OEMs, distributors, suppliers (SMEs as well as large companies), NGOs and non-profit organizations, associations, interest groups, regulators and research institutes. In the end, out of the performed literature review, the screening of the sustainability reports, the qualitative interviews and the quantitative survey a comprehensive list of 36 relevant sustainability aspects for the electronics and automotive supply chains could be identified and prioritized. The main results for this part were
- A structured list of sustainability dimensions, aspects and indicators within the supply-chain.
- A ranking of the identified sustainability topics setting the focus for further processing in the SustainHub project.
For the definition of the required environmental, social and managerial raw data, a combined method of literature review, expert interviews and expert judgements were applied. For the basic list of 115 sustainability aspects, which were compiled in the literature review, it was defined whether the aspect could be covered with a quantitative, a qualitative or both types of indicators. An online investigation for supply chain indicators led to the assumption that many sets of indicators are already available, but none of them takes into account all three dimensions of sustainability. Neither do they consider a life cycle oriented approach, with sustainability perspective, focused on the target industries. Therefore, a new set of indicators was developed for the Eco-/Sustainability Network of SustainHub.
These documents all provide sets of indicators, which are strongly varying in their level of detail, their comprehensibility from a sustainability perspective, their practicability for supply chain specific assessment and their possibility for aggregation. As it is acknowledged in literature, the consideration of the social dimension of sustainability in supply chains and therefore the development of social indicators poses a big challenge.
Hence, this shortcoming in sustainability assessment was addressed in particular, since so far there was no summable set of indicators available. Since a further aim of SustainHub is to serve as decision support instrument on all levels of the supply chain, also a managerial perspective was taken into account. In order to identify the data requirements, a table was established to define diversity and format of the data. The main results for this part were
- For each of the 36 identified aspects (see T3.1) raw data requirements from the environmental, social and managerial dimensions as well as corresponding indicators are elaborated (“data descriptions”).
- The data granularity for this specific context is defined.
- The internal data collection process within SustainHub is defined.
- A procedure for identifying sustainability aspects for SustainHub in the future is defined.
The development of the universal data model has started by a research of standards within sustainability data exchange (standardized templates). The goal was to indicate which sustainability topics are already supported by data exchange standards or quasi standards. Standardized data entities were directly incorporated into the data model whereas for quasi standards and non-standardized topics according data entities were prepared. All entities were analyzed if they belong to accompanying / general information, common to all data exchange processes or if they are specific to certain sustainability topics. Additionally, based on the preceding deliverables data model requirements have been deducted which need to be implemented to assure plausibility checks as well as the data model to be flexible for future sustainability topics. Hence, universal data model has been structured into different levels according to the re-usability of data (common / general data, background and specific data). For the moment the data model is in the revision process. The data exchange itself is based on the idea of a “general envelope” in which depending on the specific topic certain “contents” will be entered. For this an encryption and decryption mechanism has been described. The main results for this part were
- A draft of universal data model.
- A description of the encryption and decryption mechanism.
- A superior model for data exchange is elaborated.
The framework for the sustainability data was further detailed and documented. An analysis of the maturity of data exchange and standardization per indicator was concluded. A concept for the format transformation between indicators was specified, the security mechanism was refined and results were documented in D3.2. In a first step a basic structure (data categories) was elaborated and proposed for the data model. For the universal (not-indicator specific) data categories (e.g. requester) detailed data entities have been defined. For each data entity it was checked if there is equivalence in established data exchange standards (official standards as well as established / implemented industrial standards). Deviations and correspondences were described and documented. Finally the hazardous substance indicator (e.g. RoHS declaration) was described in depth along with suitable data entities. The compatibility with various international standards (e.g. Global Reporting Initiative – latest version “G4”, ISO 50.001 OHSAS, LCA etc.) was insured and supplemented with indicator-specific data fields for the data exchange systems, with a focus on managerial, ethical and social indicators. The data fields for the sustainability indicators were analyzed and incorporated into the data model concept. The format transformation was specified and documented.
A step-wise approach was used for developing the concept for plausibility checks of eco-/sustainability information. The identified problems were assessed on the relevance for SustainHub and finally grouped into four categories data measurement, data collection, data calculation and data entry. Based on this, a list of data validation checks was created and a framework for background information checking designed. The linkage of the causes for implausible data entries with the developed plausibility checks, as well as the definition of plausibility rules was established. Based on this, it was described how the concept can be applied to the indicators developed in the preceding work, by illustrating which type of indicators require which type of plausibility checks. Finally, it was outlined, how the concept for plausibility checks could be also used for company internal and external benchmarking. The main results for this part were
- A concept for plausibility checks from a professional perspective.
- A set of plausibility rules.
In order to use sustainability data to steer activities within the supply chain, it is necessary that the gathered data are aggregated. This task started by analyzing the established indicators for their capability to be aggregated as well as indicating for which sustainability area an aggregation would make sense (e.g. CO2 emissions versus gender data). Based on this a concept has been developed for data surveying and weighing in order to guarantee the interpretability. The main results for this part were
- A concept for aggregation of sustainability data / indicators.
- A definition of requirements and challenges for the aggregation method for a supply chain.
To support the identification of sustainable data an Eco-Wizard was developed. Necessary elements and the user interface were designed and sample content for the wizard created. The content creation process was documented for the example of gender aspects with workflows and written manuals and step-by-step-instructions. The wizard contents for the other defined use cases were also developed and integrated into the software prototype. The main results for this part were
- A definition of necessary elements for an eco-wizard.
- A documented example of an eco-wizard for a certain indicator.
The Code of Conduct for SustainHub was developed by analyzing relevant and already existing Codes of Conduct and common strategies at the industrial partners. Accompanying this interviews with decision makers from the industrial partners and analyzing additional company documents (reports) helped to get a better impression of the company needs. Based on this a concept for the integration of sustainability in general and SustainHub specifically to support corporate decision making in the supply chain, corporate strategy and marketing was developed and verified with all project partners. The main results for this part were
- A code of conduct for SustainHub.
S/T VI: Collection and managing eco-/sustainability data in the supply chain. To provide this function, techniques for mining and analyzing sustainability data collected from different data sources are needed.
This task started with conducting a comprehensive the state-of-the-art analysis, revealing that SustainHub’s ambitions extend well beyond the scope and technology of contemporary methods and solutions. Each of the four analysis perspectives - data extraction, data collection processes, data exchange hubs and data mining evidence a gap between the current state of the art and the planned SustainHub. Data extraction has been identified as a domain with high applicability, but low adoption. Few methods and tools possess the sophisticated capabilities that are necessary for SustainHub’s implementation. This is a niche of information technology which has so far found little application in compliance and sustainability data exchange. For the purpose of connecting between in-house systems, special requirements with respect to formats and standards have been identified.
The analysis of collection processes revealed that the advanced process management technologies have so far been underutilized in the market. A range of scientific approaches has been analyzed, none of which appears directly applicable within SustainHub. In order to fulfil the user requirements regarding process support, several methods that enable the management of process variant, relationships between data and users to processes, as well as run-time flexibility , will need to be combined and enhanced. So far, no commercial software tool exists that is coming close to producing such capabilities. In contrast with these results, Data Exchange Hubs have been found to be a popular application in compliance and sustainability data exchange. However, an in-depth analysis of the technical architecture of the market solutions against an evaluation framework evidences that most of them are relatively isolated platforms with limited integration capabilities. In particular, there are no solutions on the market which allow inter systems communication without a central data storage. Similarly, none of the market solutions covers the breadth of indicators and data types which are in the focus of SustainHub. The state of the art in data mining is a wide field of highly specialized applications. A number of exemplary data mining use cases in the context of SustainHub have been examined, together with applicable methods. Furthermore, a set of criteria has been developed, against which a selection of software vendors have been evaluated. The findings have been transformed into requirements and objectives for SustainHub’s technical solution. After that, an in-depth research on the topic of data extraction was conducted. Based on the findings of D4.1 concepts for algorithms and services for extracting sustainability data from both web sources as well as in-house systems were developed. Technical specifications for the software architecture regarding data acquisition and extraction and data formats as XSD stubs for data exchange have been developed. The main results for this part were
- A state-of-the-art report on technical solutions for collecting and managing data in a supply chain.
- A list of technical specifications for the technical solution / definition of algorithms and services (data mining algorithms, data extraction from heterogeneous sources, data collection process).
Concepts for flexible and adaptive data collection processes have been developed. A meta model for configurable process templates has been designed. Furthermore, an approach for enabling the dynamic generation, configuration, monitoring and dynamic adaptation of the data collection processes was defined. Special attention has been devoted to taking into account exception handling techniques. It has been shown how these collection processes can be implemented using an adaptive process engine. These results have been described as part of D4.2.
In the process of defining an approach matching the requirements of SustainHub, a comprehensive review of existing approaches was conducted. As automated and dynamic data collection in such a complex and heterogeneous environment is a very specific challenge, no advanced industrial solutions could be found satisfying these exact requirements. Therefore a number of scientific approaches targeting different related areas were taken into account, for example, automated support of supply chain communication, automated complex data collection, and various approaches for integrating dynamicity into process execution. On basis of this newly gathered knowledge, a number of technical requirements have been elicited. Based on this a meta model and algorithms for applying dynamic processes for complex sustainability data collection tasks in a supply chain was developed. This comprises incorporating and modelling all relevant factors that could have an influence on the structure of the process. Furthermore, an approach was defined, enabling the dynamic generation/configuration of these processes due to the modelled parameters, the monitoring of the processes as well as the influential parameters, and the dynamic adaptation of the processes to match changing needs of the current situation. The main results for this part were
- Technical requirements for the data collection processes are specified.
- A meta model for configurable process templates defining the basic steps for collecting and processing is established.
- Algorithms for the product-driven configuration of the data collection processes (configurable process templates) are defined.
- Monitoring services for the process are incorporated.
As this is a prototype of a state-of-the-art research project, a set of important functionalities of the developed concepts has been implemented within this prototype. Thereby, the partners aimed at satisfying the key challenges and requirements defined by the description of work respectively important use cases in close cooperation with the industrial partners. Another very relevant task was the preparation of the developed concepts for their application. This task comprised concretizing the concepts in order to implement them in the prototype. The concepts, which are not directly used in the prototype, have been generalized in order to prepare them for future use in (commercialized versions of) SustainHub as well as the scientific community.
Additionally various complex concepts were included into SustainHub for the incorporating of context knowledge as well as the application of this knowledge to enable an automated configuration of data collection processes. Another aspect concretized was the modelling of such concepts to enable SustainHub users to create custom configurations including complex error checking mechanisms in order to avoid erroneous configurations. The demonstrator is conceptualized based on the technologies from the Process-Aware Information System (PAIS) to the GUI components.
The architecture for the data exchange technology, the backbone of SustainHub has been developed. This includes a concept for processing data requests from data collection processes and an architecture that supports SustainHubs planned capabilities of memorizing links and other such techniques. The main results for this part were
- A concept for processing data requests from data collection processes.
- The overall architecture is defined and capable of supporting techniques for memorizing links to identified data sources and to maintain related information.
A prototype for the technical solution (based on the overall architecture) of SustainHub was developed and deployed with industry partners and further modified to make sure it meets the needs for test cases and industry considerations. In particular, the prototypical implementation of an in-house connection was implemented in the form of an interface with the iPoint Compliance Agent and then tested.
The data mining techniques that could be applied to SustainHub data have been specified on the basis of the three test cases. Algorithms as well as a concept for the integration of the data mining software within the hub architecture have been developed. From the findings, conclusions were drawn for the requirements of data mining technology which will be used for plausibility checks in the technical solution of SustainHub. An implementation plan for each of the use cases has been formulated. A detailed study of the integration between the Sustain Hub server and the data mining system was performed. The integration framework is decided and steps required for the integration are documented. A concept for accessing and using gathered data delivered by dynamic data collection workflows is defined. Therefore, hooks have been defined as well as data integration procedures. The main results for this part were
- Techniques and algorithms for mining and analyzing (checking plausibility of) sustainability data in the context of collection processes are described.
- Integration concept (architecture) for the data mining system is described.
A web service has been developed to integrate the software within the SustainHub architecture. Interfaces have been defined in terms of an XSD schema to which request and responses of the web service adhere using domain vocabulary from the specific indicator types.
Potential Impact:
In the end SustainHub will provide a hub which acts like a marketplace open to all services related to the processing of sustainability data. Basic functionalities like the data exchange will enable companies to exchange data on sustainability along and across the supply chain directly. Therefore each company stays in full control of sensitive data. In addition, the underlying flexible data model and the open architecture will assure that new sustainability data exchange topics can be added when required.
As the platform will not be tied to legacy systems or data formats, the system and data architecture will be fully adaptable to changing requirements. Additional automated extraction and conversion processes will render it possible that companies do not need to care about the format or file type in which they like to exchange data. This will decrease the effort and costs for data exchange in the whole supply chain significantly.
Further services on the hub will facilitate an early information acquisition on sustainability requirements, enabling companies to integrate them contemporarily into the product development.
Finally, by allowing companies to use test results collaboratively, the costs for material testing will be lowered and companies are enabled by the risk assessment tool to measure their financial risks growing out of non-compliant products.
SustainHub - when in operation - will improve the current situation regarding information about sustainable indicators to a great extent, of course starting with the handling of hazardous substances and the relevant regulations.
The dissemination activities concentrated mainly on the industry (of course presentations were given on conferences), because the objective of SustainHub was to improve the sustainable information handling along supply chains. The intention was to make the industry more aware of the problem and the SustainHub solution. This included fairs and conferences with industry participation as well as industrial newsletters and publications.
Partner iPoint plans to exploit the SustainHub results by beeing the first to operate this platform - at the moment it is planned to go into operation 2017.This partner will use his well established customer base for his product to handle conflict minerals to promote the SustainHub platform when the development towards a ready-to-use product has finished.
Partner rapidMiner will bring his solutions into this platform as a service to enhance the situation about imcomplete or missing data as well as supplying the SustainHub user with an early warning system about upcoming regulatory issues.
Partner FhG-IPA will use the developed data model and tool to support his consulting in the areas of RoHS, REACh and more and will continue the development of the risk analyzis methods as a service of the projected SustainHub platform.
The university reseach institutions will continue their research on the topics of exchanging data along supply chains, aggregating sustainability indicators and data collection processes.
List of Websites:
http://sustainhub-research.eu/
A number of new challenges for companies in the EU emerge in a globalised economic cycle.
Manufacturers and their suppliers from the electronics and automotive industries are required to comply with restrictions for selling and exporting products in the EU.
Directives like the RoHS and ELV interdict the use of certain substances. In addition, the REACH legislation requires companies to conform to a whole range of requirements related to production, import and use of materials.
Adding to this is the need to comply to the increasing wish for sustainable products / production, be it by own choice or society pressure. The reality in industrial production and trade is way behind this. Main causes for this are the lack of trustful information about sustainable indicators between suppliers and customers (caused by a missing or incomplete data exchange or missing standardization) and the insufficient knowledge, especially in smaller companies, about all sustainable issues, be it regulations or laws or possible threats emerging from wrong conduct of these laws and regulations.
SustainHub therefore combined organisational and institutional measures with technical solutions.
First of all SustainHub addressed an important factor which is a bare necessity to implement a lasting new service, a business model, which is as well stable as scalable. This business model points out a way to implement the SustainHub data exchange as a self-financed service. For this service SustainHub defined a set of valid sustainability indicators (and methods to calculate and aggregate them) and created the necessary IT tools, namely advanced communication facilities for supply chain information interchange, a data model capable of managing all aspects of product and related sustainability data, tools to expand the available information basis by data mining principles and tools to improve the data quality by using algorithms for automatically checking the plausibility of data.
Project Context and Objectives:
The Challenge
European companies are increasingly under pressure to bring sustainable products to market. Especially high technology industries such as electronics and automotive are driven not only by environmental legislation but by consumer demand to produce sustainable products. Companies trying to address this situation face an information problem: exchanging compliance and sustainability data on corporate, component and material levels across the entire supply chains. The prevailing situation in the industry is that participants take a reactive position regarding compliance and sustainability data exchange.
The motivation for OEMs is to avoid costly product recalls and public relations crises. Therefore they pass the data exchange requirements on to their supply chain, and their market power forces suppliers to comply. Suppliers in the middle / lower supply chain usually have smaller company sizes and they are under pressure to meet the data requirements, in order to assure their market access. Their role is especially resource-intensive, as they have to gather data from several other tiers and process them into required, often custom reporting formats.
In consequence those companies, which often lack dedicated resources, suitable processes and IT systems for sustainability data management, are struggling to fulfil the customers’ demands.
As a result are the current processes concerned with sustainability data processing inefficient and largely manual and necessary data is either incomplete, failure prone as well as stored in a multiplicity of formats and sources. The integration into operational internal processes is insufficient, exchange processes are slow and complicated.
Project Objectives
In the long run SustainHub will provide an open and self-augmenting collaborative marketplace for various sustainability data processing topics. Besides the overall concept for the SustainHub marketplace, the project planned to develop first fundamental services to be offered on the platform. Those services are dedicated either to increase data quality, risk and information transparency or to lower manual effort and therefore overall costs associated with the processing of sustainability data.
Regarding data exchange the objective was the development of an easy exchange solution of eco-/sustainability data along and across supply chains. For this purpose the SustainHub project had to define organizational measures like procedures to gather and aggregate sustainability data and furthermore had to establish an IT-supported collection and transformation (e.g. from one standard to another) process. The link to present databases should furthermore enable the reuse of existent data. To further support data collection within the companies, a support function had to be setup to guide inexperienced users through the collection process.
To address data actuality, correctness and quality issues, the projects objective was to provide first services which enable potential users to receive early information on necessary data as well as to check generated data.
For this, the SustainHub project planned to deliver a tool which enables the user to identify upcoming sustainability requirements in a very early stage. This shall allow companies to integrate new requirements in an early phase of the product development.
Above that, plausibility procedures are to be obtainable, which evaluate the sustainability data quality. These plausibility checks are to be supported both by checks from a professional (human) perspective as well as by automated checks based on data mining techniques. Those techniques will additionally be used to estimate data gaps and to establish an automated risk assessment on in-house sustainability data.
A collaborative testing tool had to be developed, which enables companies to share sustainability tests (e.g. XRF-tests for RoHS-Screenings).
Finally, the development work is encompassed and guided by a Business Model which shall ensure the future market access.
Project Results:
The scientific and technological results of SustainHub are:
S/T I: Collaborative Sustainability Network along the entire supply chain in order to enable the easy exchange of eco-/sustainability information along the supply chain. This means, the development of a business model for a self-augmenting collaborative network for sustainability data.
A comprehensive survey of the industrial requirements was conducted. The gathered data / requirements have been analysed in absolute numbers as well as dependent on the priority in each customer segment. Findings of the quantitative and qualitative surveys were documented in D1.1 and first conclusions for the implementation of the SustainHub platform were drawn. This included the definition of preferred functionalities and features. The main results for this part were
- Customer expectations regarding all steps of sustainability data processing from collection through data preparation and analysis to data transfer. Each result has been specified according to tier position and company size.
- Importance of different sustainability topics and their relevance for data exchange in the supply chain and the efforts linked to the data processing identified.
- Information regarding customer segmentation, incentive and pricing strategy available
The Identified relevant sustainability issues were analyzed for their complexity. Issues like type (qualitative or quantitative) and format of data, completeness, transparency and actuality of data as well as external verification of data were considered. In order to test and validate results, experts with profound experience in supply chain data exchange in the automotive and electronics industries were consulted to test and assess the organizational data requirements. Finally the input of experts in the field of sustainability management was taken into account to select the most appropriate type, diversity and amount of data for the indicators. Lastly the requirements were evaluated and defined from an organizational and technical point of view. A case example of a manual, which describes the presentation of quantitative and qualitative data, was established. For this purpose gender and energy related data were selected. The main results for this part were
- Requirements arising from sustainability data are defined (complexity in terms of diversity, type, amount).
- Necessary sustainability raw data is defined as a basis for the development of the Data Model.
- Manual for the identification of environmental and Sustainable raw data.
- 36 sustainability indicators are selected (out of 115 aspects) for further processing in the SustainHub project.
Based on the output of different workshop sessions and research done by the project partners a common vision on the business model has been defined. Different alternative options and solutions have been formulated and prioritized. This makes the business model flexible towards future changes. Due to interdependencies with the technical requirements, several specifications of the business model have been adapted in several feedback loops. The main results for this part were
- A Business framework and a complete roadmap for the successful introduction of SustainHub into the market is concluded.
- The Customer segmentation and key players in the field of concern are defined.
- Overall value proposition of the marketplace and the specific services by addressing each customer segment is specified.
- Necessary resources are calculated.
- A financial model (pricing and costs) is established.
- Marketing strategies are deducted.
- An incentive model to guarantee high participation (with special consideration of SMEs) is specified.
The requirements for the SustainHub business model and for the technical concepts have been consolidated from the requirements and documented in D1.4. The technical requirements have been transformed into user stories and prioritized, in order to scope the prototype implementation. From this a detailed list of technical use cases with implementation specifications has been elicited from the user stories. This specification sheet has been designed in such a way that it could be utilized as guidance for the development of a prototype. The developed business model has been evaluated with consortium partners and interested third parties. The main results for this part were
- Consolidated requirements with impact on the technical solution.
- Prioritized user stories for technical solution.
- A specification of technical uses cases for the implementation.
- A roadmap for the technical development.
- Evaluated assumptions (by potential industrial customers ) of the business model and the technical concept.
S/T II: Scouting tool for eco- and sustainability requirements for their rapid integration in the design process of products and the avoidance of expensive redesign cycles. To reach this objective, reliable tools for the early detection and recognition of changing or new eco-/sustainability requirements are necessary.
A list of product characteristics with significant impact on the eco-efficiency of products has been gathered from various guidelines and articles. This list has been structured and a general target and lever system for eco-efficiency from a product perspective has been developed. This can be easily adopted in product development and for example in the customer requirements survey. Where possible the list was supplemented by a set of international standards for each characteristic. Based on the presented system a methodology was developed which describes – on an aggregated level – how certain upcoming requirements related to the defined product characteristics could be detected precociously on the internet. Defining the according sentiments linked the model to the technical implementation of the Eco-scouting tool. The main results for this part were
- A list of product characteristics with impact on the eco-efficiency of products. This list gives a well-defined orientation on what kind of characteristics need to be taken into account in product development.
- A procedure how to precociously detect requirements regarding eco-efficiency of products on the internet.
- A list of sentiments.
- An eco-scouting tool – search and analysis engine for material related product characteristics (e.g. Hazardous Substances).
S/T III: Collaborative Testing Network in order to save resources for testing components and materials (e.g. RoHS issues) a common use and exchange of the results is planned. This network will reduce the number of needed tests and efforts for every single member while the data quality is improved. The benefits of using the testing network are they will strengthen the dissemination of SustainHub.
The base for the Collaborative Testing Network (CTN) is the need for reliable storage and retrieval of test data. This includes the mapping of test results to components or materials. A relational database was concluded by comprising of all necessary data entities (demographic data plus BoM and all risk influencing parameters). Especially the intelligent testing procedures, material classifications, material properties, categorization of legal requirements and chemical elements have been defined and prepared according to the database design. Several thousand real data sets have been imported repeatedly and tools for the import from other databases have been established. Based on the developed use cases, rules for the risk assessment are described and have partly been programmed. A concept for the organization of collaborative testing has been created. This allows the participating companies in the network to decrease their operational risks of non-compliance and also reduce their efforts and expenditure on testing. The developed data base and user interface has the potential to help companies to ensure the fulfilment of their due diligence as required by the regulation (EC) No 765/2008 (market surveillance). The main results for this part were
- A concept for intelligent testing.
- A model of risk influencing parameters.
- A relational database is operational, import tools are programmed, the database is filled with necessary data (regulations, substances, examples for BoM).
- A concept for risk calculation is defined and can be demonstrated.
- A concept for the organization of collaborative testing is defined.
One aspect of the business model mentioned above was the future integration of further service partners into the SustainHub. In a first step different kinds of possible future service providers were defined (e.g. content providers, service providers like consultants). This resulted in a general qualification of possible future services along the steps of data processing. Furthermore the integration process was described in terms of quality management using a QFD driven approach for requirements analysis and revision processes. Finally, an according business and incentive model for future service providers was added in order to lay down how they could be attracted to the SustainHub platform. The main results for this part were
- A method is described how additional services can be integrated into the platform.
- The qualification of services is defined.
- Quality management procedures are set up for the integration process.
- A business and incentive model for additional service providers is defined.
S/T IV: Methods for validation and estimation of missing eco-/sustainability data, within this objective the development of methods for validation of existing and the estimation of missing data will be realized. Using these methods will improve the quality of the data and enhance the acceptance of the data network.
The areas in the database, in which data could be missing / be false and what kind of data types could be affected in what way were defined. A list of queries for the data estimation / validation was created which shows the different problems and goals regarding missing / false data. Based on real data sets different validation and estimation strategies have been developed. In summary, the data mining algorithms could be successfully applied to achieve the use cases for the SustainHub platform. Supervised classification techniques outperform the association mining techniques and also supervised classification techniques could be successfully used to achieve the use cases for missing data and plausibility checks. Though there were challenges due to limited data access, the results obtained for the substances that are available in the prototypical data look promising. With the availability of more data, we expect the system could perform better and the reliability on the results from data mining can significantly improve. The main results for this part were
- A list of use cases / queries for the data mining.
- Different mining strategies and procedures are applied and evaluated.
- Several data patterns are identified and documented.
S/T V: Identification and definition of environmental and sustainability data. A concept for the quick and easy integration of changing or new requirements concerning eco-/sustainability aspects in the supply chain will be needed. Based on case examples, a code of conduct will be determined and defined.
The sustainability aspects within the supply chain were evaluated through a comprehensive literature research, the evaluation of numerous sustainability reports and by interviews with industrial partners. A deductive quantitative and qualitative content analysis was applied on the interviews. The deductive approach was used as a ranking criterion for identification of the most relevant sustainability aspects. Due to the combination of both approaches, more reliable and significant results were achieved. Finally, a quantitative online survey was designed in order to serve as an additional filter for the identified sustainability aspects. The survey also helped to better understand the relevance of sustainable supply chains for various stakeholders and was available in English, German and French. The aim was to gather approximately 100 answers from OEMs, distributors, suppliers (SMEs as well as large companies), NGOs and non-profit organizations, associations, interest groups, regulators and research institutes. In the end, out of the performed literature review, the screening of the sustainability reports, the qualitative interviews and the quantitative survey a comprehensive list of 36 relevant sustainability aspects for the electronics and automotive supply chains could be identified and prioritized. The main results for this part were
- A structured list of sustainability dimensions, aspects and indicators within the supply-chain.
- A ranking of the identified sustainability topics setting the focus for further processing in the SustainHub project.
For the definition of the required environmental, social and managerial raw data, a combined method of literature review, expert interviews and expert judgements were applied. For the basic list of 115 sustainability aspects, which were compiled in the literature review, it was defined whether the aspect could be covered with a quantitative, a qualitative or both types of indicators. An online investigation for supply chain indicators led to the assumption that many sets of indicators are already available, but none of them takes into account all three dimensions of sustainability. Neither do they consider a life cycle oriented approach, with sustainability perspective, focused on the target industries. Therefore, a new set of indicators was developed for the Eco-/Sustainability Network of SustainHub.
These documents all provide sets of indicators, which are strongly varying in their level of detail, their comprehensibility from a sustainability perspective, their practicability for supply chain specific assessment and their possibility for aggregation. As it is acknowledged in literature, the consideration of the social dimension of sustainability in supply chains and therefore the development of social indicators poses a big challenge.
Hence, this shortcoming in sustainability assessment was addressed in particular, since so far there was no summable set of indicators available. Since a further aim of SustainHub is to serve as decision support instrument on all levels of the supply chain, also a managerial perspective was taken into account. In order to identify the data requirements, a table was established to define diversity and format of the data. The main results for this part were
- For each of the 36 identified aspects (see T3.1) raw data requirements from the environmental, social and managerial dimensions as well as corresponding indicators are elaborated (“data descriptions”).
- The data granularity for this specific context is defined.
- The internal data collection process within SustainHub is defined.
- A procedure for identifying sustainability aspects for SustainHub in the future is defined.
The development of the universal data model has started by a research of standards within sustainability data exchange (standardized templates). The goal was to indicate which sustainability topics are already supported by data exchange standards or quasi standards. Standardized data entities were directly incorporated into the data model whereas for quasi standards and non-standardized topics according data entities were prepared. All entities were analyzed if they belong to accompanying / general information, common to all data exchange processes or if they are specific to certain sustainability topics. Additionally, based on the preceding deliverables data model requirements have been deducted which need to be implemented to assure plausibility checks as well as the data model to be flexible for future sustainability topics. Hence, universal data model has been structured into different levels according to the re-usability of data (common / general data, background and specific data). For the moment the data model is in the revision process. The data exchange itself is based on the idea of a “general envelope” in which depending on the specific topic certain “contents” will be entered. For this an encryption and decryption mechanism has been described. The main results for this part were
- A draft of universal data model.
- A description of the encryption and decryption mechanism.
- A superior model for data exchange is elaborated.
The framework for the sustainability data was further detailed and documented. An analysis of the maturity of data exchange and standardization per indicator was concluded. A concept for the format transformation between indicators was specified, the security mechanism was refined and results were documented in D3.2. In a first step a basic structure (data categories) was elaborated and proposed for the data model. For the universal (not-indicator specific) data categories (e.g. requester) detailed data entities have been defined. For each data entity it was checked if there is equivalence in established data exchange standards (official standards as well as established / implemented industrial standards). Deviations and correspondences were described and documented. Finally the hazardous substance indicator (e.g. RoHS declaration) was described in depth along with suitable data entities. The compatibility with various international standards (e.g. Global Reporting Initiative – latest version “G4”, ISO 50.001 OHSAS, LCA etc.) was insured and supplemented with indicator-specific data fields for the data exchange systems, with a focus on managerial, ethical and social indicators. The data fields for the sustainability indicators were analyzed and incorporated into the data model concept. The format transformation was specified and documented.
A step-wise approach was used for developing the concept for plausibility checks of eco-/sustainability information. The identified problems were assessed on the relevance for SustainHub and finally grouped into four categories data measurement, data collection, data calculation and data entry. Based on this, a list of data validation checks was created and a framework for background information checking designed. The linkage of the causes for implausible data entries with the developed plausibility checks, as well as the definition of plausibility rules was established. Based on this, it was described how the concept can be applied to the indicators developed in the preceding work, by illustrating which type of indicators require which type of plausibility checks. Finally, it was outlined, how the concept for plausibility checks could be also used for company internal and external benchmarking. The main results for this part were
- A concept for plausibility checks from a professional perspective.
- A set of plausibility rules.
In order to use sustainability data to steer activities within the supply chain, it is necessary that the gathered data are aggregated. This task started by analyzing the established indicators for their capability to be aggregated as well as indicating for which sustainability area an aggregation would make sense (e.g. CO2 emissions versus gender data). Based on this a concept has been developed for data surveying and weighing in order to guarantee the interpretability. The main results for this part were
- A concept for aggregation of sustainability data / indicators.
- A definition of requirements and challenges for the aggregation method for a supply chain.
To support the identification of sustainable data an Eco-Wizard was developed. Necessary elements and the user interface were designed and sample content for the wizard created. The content creation process was documented for the example of gender aspects with workflows and written manuals and step-by-step-instructions. The wizard contents for the other defined use cases were also developed and integrated into the software prototype. The main results for this part were
- A definition of necessary elements for an eco-wizard.
- A documented example of an eco-wizard for a certain indicator.
The Code of Conduct for SustainHub was developed by analyzing relevant and already existing Codes of Conduct and common strategies at the industrial partners. Accompanying this interviews with decision makers from the industrial partners and analyzing additional company documents (reports) helped to get a better impression of the company needs. Based on this a concept for the integration of sustainability in general and SustainHub specifically to support corporate decision making in the supply chain, corporate strategy and marketing was developed and verified with all project partners. The main results for this part were
- A code of conduct for SustainHub.
S/T VI: Collection and managing eco-/sustainability data in the supply chain. To provide this function, techniques for mining and analyzing sustainability data collected from different data sources are needed.
This task started with conducting a comprehensive the state-of-the-art analysis, revealing that SustainHub’s ambitions extend well beyond the scope and technology of contemporary methods and solutions. Each of the four analysis perspectives - data extraction, data collection processes, data exchange hubs and data mining evidence a gap between the current state of the art and the planned SustainHub. Data extraction has been identified as a domain with high applicability, but low adoption. Few methods and tools possess the sophisticated capabilities that are necessary for SustainHub’s implementation. This is a niche of information technology which has so far found little application in compliance and sustainability data exchange. For the purpose of connecting between in-house systems, special requirements with respect to formats and standards have been identified.
The analysis of collection processes revealed that the advanced process management technologies have so far been underutilized in the market. A range of scientific approaches has been analyzed, none of which appears directly applicable within SustainHub. In order to fulfil the user requirements regarding process support, several methods that enable the management of process variant, relationships between data and users to processes, as well as run-time flexibility , will need to be combined and enhanced. So far, no commercial software tool exists that is coming close to producing such capabilities. In contrast with these results, Data Exchange Hubs have been found to be a popular application in compliance and sustainability data exchange. However, an in-depth analysis of the technical architecture of the market solutions against an evaluation framework evidences that most of them are relatively isolated platforms with limited integration capabilities. In particular, there are no solutions on the market which allow inter systems communication without a central data storage. Similarly, none of the market solutions covers the breadth of indicators and data types which are in the focus of SustainHub. The state of the art in data mining is a wide field of highly specialized applications. A number of exemplary data mining use cases in the context of SustainHub have been examined, together with applicable methods. Furthermore, a set of criteria has been developed, against which a selection of software vendors have been evaluated. The findings have been transformed into requirements and objectives for SustainHub’s technical solution. After that, an in-depth research on the topic of data extraction was conducted. Based on the findings of D4.1 concepts for algorithms and services for extracting sustainability data from both web sources as well as in-house systems were developed. Technical specifications for the software architecture regarding data acquisition and extraction and data formats as XSD stubs for data exchange have been developed. The main results for this part were
- A state-of-the-art report on technical solutions for collecting and managing data in a supply chain.
- A list of technical specifications for the technical solution / definition of algorithms and services (data mining algorithms, data extraction from heterogeneous sources, data collection process).
Concepts for flexible and adaptive data collection processes have been developed. A meta model for configurable process templates has been designed. Furthermore, an approach for enabling the dynamic generation, configuration, monitoring and dynamic adaptation of the data collection processes was defined. Special attention has been devoted to taking into account exception handling techniques. It has been shown how these collection processes can be implemented using an adaptive process engine. These results have been described as part of D4.2.
In the process of defining an approach matching the requirements of SustainHub, a comprehensive review of existing approaches was conducted. As automated and dynamic data collection in such a complex and heterogeneous environment is a very specific challenge, no advanced industrial solutions could be found satisfying these exact requirements. Therefore a number of scientific approaches targeting different related areas were taken into account, for example, automated support of supply chain communication, automated complex data collection, and various approaches for integrating dynamicity into process execution. On basis of this newly gathered knowledge, a number of technical requirements have been elicited. Based on this a meta model and algorithms for applying dynamic processes for complex sustainability data collection tasks in a supply chain was developed. This comprises incorporating and modelling all relevant factors that could have an influence on the structure of the process. Furthermore, an approach was defined, enabling the dynamic generation/configuration of these processes due to the modelled parameters, the monitoring of the processes as well as the influential parameters, and the dynamic adaptation of the processes to match changing needs of the current situation. The main results for this part were
- Technical requirements for the data collection processes are specified.
- A meta model for configurable process templates defining the basic steps for collecting and processing is established.
- Algorithms for the product-driven configuration of the data collection processes (configurable process templates) are defined.
- Monitoring services for the process are incorporated.
As this is a prototype of a state-of-the-art research project, a set of important functionalities of the developed concepts has been implemented within this prototype. Thereby, the partners aimed at satisfying the key challenges and requirements defined by the description of work respectively important use cases in close cooperation with the industrial partners. Another very relevant task was the preparation of the developed concepts for their application. This task comprised concretizing the concepts in order to implement them in the prototype. The concepts, which are not directly used in the prototype, have been generalized in order to prepare them for future use in (commercialized versions of) SustainHub as well as the scientific community.
Additionally various complex concepts were included into SustainHub for the incorporating of context knowledge as well as the application of this knowledge to enable an automated configuration of data collection processes. Another aspect concretized was the modelling of such concepts to enable SustainHub users to create custom configurations including complex error checking mechanisms in order to avoid erroneous configurations. The demonstrator is conceptualized based on the technologies from the Process-Aware Information System (PAIS) to the GUI components.
The architecture for the data exchange technology, the backbone of SustainHub has been developed. This includes a concept for processing data requests from data collection processes and an architecture that supports SustainHubs planned capabilities of memorizing links and other such techniques. The main results for this part were
- A concept for processing data requests from data collection processes.
- The overall architecture is defined and capable of supporting techniques for memorizing links to identified data sources and to maintain related information.
A prototype for the technical solution (based on the overall architecture) of SustainHub was developed and deployed with industry partners and further modified to make sure it meets the needs for test cases and industry considerations. In particular, the prototypical implementation of an in-house connection was implemented in the form of an interface with the iPoint Compliance Agent and then tested.
The data mining techniques that could be applied to SustainHub data have been specified on the basis of the three test cases. Algorithms as well as a concept for the integration of the data mining software within the hub architecture have been developed. From the findings, conclusions were drawn for the requirements of data mining technology which will be used for plausibility checks in the technical solution of SustainHub. An implementation plan for each of the use cases has been formulated. A detailed study of the integration between the Sustain Hub server and the data mining system was performed. The integration framework is decided and steps required for the integration are documented. A concept for accessing and using gathered data delivered by dynamic data collection workflows is defined. Therefore, hooks have been defined as well as data integration procedures. The main results for this part were
- Techniques and algorithms for mining and analyzing (checking plausibility of) sustainability data in the context of collection processes are described.
- Integration concept (architecture) for the data mining system is described.
A web service has been developed to integrate the software within the SustainHub architecture. Interfaces have been defined in terms of an XSD schema to which request and responses of the web service adhere using domain vocabulary from the specific indicator types.
Potential Impact:
In the end SustainHub will provide a hub which acts like a marketplace open to all services related to the processing of sustainability data. Basic functionalities like the data exchange will enable companies to exchange data on sustainability along and across the supply chain directly. Therefore each company stays in full control of sensitive data. In addition, the underlying flexible data model and the open architecture will assure that new sustainability data exchange topics can be added when required.
As the platform will not be tied to legacy systems or data formats, the system and data architecture will be fully adaptable to changing requirements. Additional automated extraction and conversion processes will render it possible that companies do not need to care about the format or file type in which they like to exchange data. This will decrease the effort and costs for data exchange in the whole supply chain significantly.
Further services on the hub will facilitate an early information acquisition on sustainability requirements, enabling companies to integrate them contemporarily into the product development.
Finally, by allowing companies to use test results collaboratively, the costs for material testing will be lowered and companies are enabled by the risk assessment tool to measure their financial risks growing out of non-compliant products.
SustainHub - when in operation - will improve the current situation regarding information about sustainable indicators to a great extent, of course starting with the handling of hazardous substances and the relevant regulations.
The dissemination activities concentrated mainly on the industry (of course presentations were given on conferences), because the objective of SustainHub was to improve the sustainable information handling along supply chains. The intention was to make the industry more aware of the problem and the SustainHub solution. This included fairs and conferences with industry participation as well as industrial newsletters and publications.
Partner iPoint plans to exploit the SustainHub results by beeing the first to operate this platform - at the moment it is planned to go into operation 2017.This partner will use his well established customer base for his product to handle conflict minerals to promote the SustainHub platform when the development towards a ready-to-use product has finished.
Partner rapidMiner will bring his solutions into this platform as a service to enhance the situation about imcomplete or missing data as well as supplying the SustainHub user with an early warning system about upcoming regulatory issues.
Partner FhG-IPA will use the developed data model and tool to support his consulting in the areas of RoHS, REACh and more and will continue the development of the risk analyzis methods as a service of the projected SustainHub platform.
The university reseach institutions will continue their research on the topics of exchanging data along supply chains, aggregating sustainability indicators and data collection processes.
List of Websites:
http://sustainhub-research.eu/