One of the most exciting new capabilities in Smart Manufacturing (SM) and Cyber-Physical Production Systems (CPPS) is the provisioning of manufacturing services as unbundled "apps or services," which could be significantly more flexible and less expensive to use than the current generation of monolithic manufacturing applications. However, bundling and integrating heterogeneous services in the form of such apps or composite services is not a trivial job. There is a need for service vendors, cloud vendors, manufacturers, and other stakeholders to work collaboratively to simplify the effort to "mix-and-match" and compose the apps or services.
This workshop is the fourth in the series of smart manufacturing workshops organized by the National Institute of Standards and Technology (NIST) and the Open Applications Group Inc. (OAGi) to address technology and standards needs for easier discovery and easier integration of services based on improved interoperability and composability. Joining the organizers again this year is the SM & CPPS Special Interest Group (SIG), recently established under the International Federation of Information Processing (IFIP) WG 5.7, Advances in Production Management Systems.
At the first Open Cloud Architecture for Smart Manufacturing workshop, the attendees recognized that open cloud service ecosystems provide a promising new platform to advance innovations in manufacturing production and supply chain management. Yet, they also identified and prioritized a number of standards and technology issues that hinder the adoption of the new innovation platform. Participants indicated the need to close the gaps and overlaps among standards, and the need to simplify and assist in the use of standards and SM technologies, as top issues.
At the second Drilling down on Smart Manufacturing -- Enabling Composable Apps workshop, the attendees started work to address the identified top issues within five working sessions. The first three working sessions were focused on the analysis, methods, and tools to address those top issues. The sessions include SM model-based standards development, Standards capability analysis for SM, and SM systems characterization. The other two sessions looked into realizing the innovation platform. The sessions include SM apps and service marketplaces and Crowdsourcing of manufacturing knowledge.
At the third Enabling Composable Service-Oriented Manufacturing Systems workshop, the attendees continued to address the identified top issues within five working sessions. Three working sessions continued with the previous year focus: SM model-based standards development, Standards capability analysis for SM, and SM apps and service marketplaces. The other two sessions adapted their focus: SM standards capability analysis became SM reference models and reference architecture; Crowdsourcing of manufacturing knowledge became Industrial Ontology Foundry.
Call for Participation:
This year's workshop will continue work within the five working sessions from the previous year. The plan is for this year to refine the identified goals, capability gaps, and needed technology descriptions for each of the sessions. Where possible, sessions will propose one or more priority roadmap topics to be addressed by the community, based on the interests and commitment from the community to work on these topics. As in the previous year, we will continue to use the common theme of reference models and their life-cycle management, as key enablers of Composable Service-Oriented Manufacturing Systems. It is expected each session will take its own perspective on this topic and lead towards creating priority roadmap topics.
The workshop will start with opening plenary where the session chairs present objectives and planned work for their sessions. A joint session will follow where cross-cutting presentations will be provided to promote cross-pollination of ideas across the breakout sessions. The workshop will continue with five parallel working sessions where the participants will discuss the existing and propose new material to define focus of the future community efforts within priority roadmap topics. Each session will have its specific focus and an agenda that will be formulated and announced by the session chairs. Please direct your interests and session-specific questions to the respective session chairs indicated below.
Interested professionals and researchers are invited to submit short statements including their positions on these workshop topics, contribute their materials, and share technical insights. Please contact the session chairs and copy the workshop chairs, indicating your intent to participate and/or submit your contributions. The results from the workshop will be published as a compendium of reports from working sessions as contributed by session chairs.
Session chairs and general descriptions are provided below. Planned work and objectives will follow.
SM model-based message standards development : Chairs – nivezic [at] nist.gov (Dr. Nenad Ivezic (link sends e-mail)), NIST and serm [at] nist.gov (Dr. Serm Kulvatunyou (link sends e-mail)), NIST. The session is seeking to advance the methodology for messaging standards (e.g., OAGIS) development and usage. The vision of the group is to develop model-driven method and tools that drive more effective and easier-to-use messaging standards. Latest developments, such as business process context-based usage and life-cycle management of messaging standards will be discussed.
SM systems characterization : Chairs – kym.wehrle [at] uilabs.org (Kym Wehrle (link sends e-mail)), DMDII and michael.brundage [at] nist.gov (Dr. Michael Brundage (link sends e-mail)), NIST. Harmonization of digital assessment tools is needed to accelerate the adoption of digital technologies within the industrial base and Department of Defense. A number of organizations are funding projects to understand their current digital state and the implications of adopting digital technology within their operations. This session will explore the activities required to harmonize common digital technology principals, approaches and tools to ultimately strengthen U.S. manufacturing competitiveness.
SM reference models and reference architecture : Chairs – yan.lu [at] nist.gov (Dr. Yan Lu (link sends e-mail)), NIST and nollerd [at] us.ibm.com (Mr. Dave Noller (link sends e-mail)), IBM. This session is intended to bring standards developers, technology providers and manufacturers together to discusses impacts of ICT technologies on the emerging manufacturing system architecture. More specifically, we would like to explore how service-oriented-architecture can help integrate IoT, digital factory and cloud computing technologies into modern manufacturing environment and enable the manufacturing systems to respond in real time to meet changing demands and conditions in the factory, in the supply network, and in customer needs. Challenges will be identified for implementing SOA based smart manufacturing systems and standards in support of such implementations will be analyzed.
SM apps and service marketplaces: Chairs – u1o [at] psu.edu (Dr. Soundar Kumara ), Penn State University and thwuest [at] mail.wvu.edu (Dr. Thorsten Wuest ), West Virginia University (WVU).In this session, the potential and current status of SM apps and service marketplaces will be discussed. The aim is to work towards shared, secure, open-access infrastructure rich in functionality for easier systems integration and composability and a marketplace that can drive technological capability beyond just products by integrating services on standards, uncertainty quantification, benchmarking, performance-use metrics, systems modeling, and many more. A special focus of this year’s workshop will be on current technological and other challenges, interoperability and security issues as well as requirements from the stakeholders' (e.g., designers, providers & users) perspectives.
Industrial Ontology Foundry: Chairs – dimitris.kiritsis [at] epfl.ch (Dr. Dimitris Kiritsis (link sends e-mail)), EPFL, evan.wallace [at] nist.gov (Mr. Evan Wallace, NIST (link sends e-mail)). The session focuses on the formation of an Industrial Ontologies Foundry (IOF), a new effort for converging existing semantic representations from the industrial and manufacturing domain. The primary purpose of the IOF is to develop a collaborative framework and platform for supporting the development, submitting, validating, and sharing ontologies for the industrial and manufacturing domains. In this way, the knowledge can be captured in a common semantic form and shared to facilitate smart manufacturing and other industrial practices and resources along the lifecycle of a manufactured product. This year’s session will review the structure of this new organization, what we’ve learned from an initial proof-of-concept effort, and the principles and processes that should be used to by the IOF to deliver value to the manufacturing industry.
Data Analytics: Chairs: Dr Al Jones, NIST (albert.jones [at] nist.gov (albert[dot]jones[at]nist[dot]gov)) and Dr. Wilawan Onkham, UPS (wonkham [at] ups.com (wonkham[at]ups[dot]com)). This session will focus on both technical and interface obstacles associated with manufacturers using cloud-based, data-analytics (DA) services. We will spend time discussing solutions to four such problems. The first involves matching specific, manufacturing problems to specific, data-analytics, solution algorithms. For example, what kinds of manufacturing problems are best solved using neural networks and how can we choose the appropriate algorithm for the available data. The second involves estimating uncertainties associated with using those algorithms; their software implementations; and any exogenous factors impacting the results. The third involves extend the existing predictive model markup language (PMML) to include standardized guidelines for helping manufacturers create the models and training data needed to use PMML. The fourth topic involves measuring the accuracy of data-analytics models in the real-world of manufacturing. This means that it is impossible to build a completely error-free DA model. This is true regardless of 1) the amount, type, and quality of the input data and 2) the complexity of the manufacturing. It goes without saying, that the ability to measure that accuracy is critical.