Projects

Example projects:

Proto-OKN: Supply and Demand Open Knowledge Network (SUDOKN) project 13

NSF AWARD 2333801: The vulnerability of the U.S manufactures and supply chains has recently become more pronounced due to various reasons such disruptions in global trade, rising labor cost, insufficient investment in manufacturing workforce, and shortage of key materials and components. Future manufacturing supply networks need to be supported by novel cyber-enabled and AI-powered platforms and data-driven solutions that can provide supply chain decision makers with real-time insight into the strengths and weaknesses of manufacturing companies and workforce, as well as the potential areas of risk and vulnerability. In this project, we will prototype and deploy the Supply and Demand Open Knowledge Network (SUDOKN) is composed of several open and interconnected knowledge graphs, aligned with formal ontologies, that collectively represent various types of supply and demand data needed to address the challenges posed by the use cases related to supplier discovery and capability and capacity analysis. Our proposal aims at democratizing access to publicly available supply and demand data maximizing its utility by proposing a three-pronged approach: (1) Develop a set of principle-based, accurately axiomatized, and reusable ontologies for modeling the domain  (2) Develop the tools required to ingest, gather, process, and search manufacturing data and (3) Use these tools and ontologies to build an open Manufacturing Capability Network (MCN) and its supporting knowledge graphs from diverse datasets .

A Hybrid Approach for Developing, Extending, and Implementing Industrial Maintenance Knowledge Graphs and Semantic Ontologies to Support Smart Maintenance Diagnostics project 12

The objective of this research is to help advance the progression from data to information and knowledge through developing Open Knowledge Network (OKN) in maintenance domain through data-driven creation of a public and open-source knowledge graph based on SKOS standard and creation of a maintenance ontology based on top-down and bottom-up approaches. In particular, this work is aimed at introducing a hybrid methodology for generating structured and machine-actionable knowledge models from the unstructured maintenance log data. The methodology uses text analytics techniques, in combination with a human-in-the-loop (HITL) thesaurus development method, for the purpose of generating a formal thesaurus (i.e., knowledge graph).  The resulting knowledge graph is intended to encode the semantic and lexical relationships between various entities in the maintenance domain. A knowledge graph allows stored data (both structured and unstructured data) to be understood at a semantic level through annotation, semantic integration and connection to  other datasets and knowledge models. The OWL ontology is used to enhance the interoperability and the semantic expressivity of the knowledge graph. The developed semantic artifacts (SKOS concept scheme and OWL ontology) will include generic entities in the upper levels of the taxonomy. Those generic entities can be applied to a wide range of maintenance applications and extended to meet specific use cases beyond the use cases and datasets that support the proposed project. The main deliverables of this projects include 1) SKOS concept scheme, 2) OWL ontology, and 3) proof-of-concept maintenance diagnostics tool and their related documentations and guides.

Ontology Modeling and Data Integration for Agri-Food Supply Chain Traceability project 11 

Traceability of food and feed is becoming an increasing concern among governments, producers, and consumers.   Governments wish to act quickly to identify and take tainted food out of the supply chain in response to a food emergency.  Producers wish to minimize their exposure to risk and ensure the quality of the food they sell.  Consumers are increasingly interested in where their food comes from, what processes were used to produce it, and what it may contain (such as pesticides or genetically modified elements).  Traceability can address all these concerns but is challenging to achieve due to the wide range of diverse participants in a supply chain spanning material source to consumer.

The Institute of Food Technologists (IFT) has proposed an approach to address some of these challenges [1].  The approach focuses on a few kinds of occurrences, which they call Critical Tracking Events (CTEs), that are key parts of the lifecycle of a product or of another participant in that product’s lifecycle.

While adopting standards for types of Critical Tracking Events, the data elements that should be captured for them, and identification schemes for related entities would address many of the current challenges for end-to-end traceability data, developing these standards and having them adopted nearly universally across food and agriculture business is both a political and practical challenge.  However, the researchers of the Supply Chain Traceability for Agri-Food project at NIST and their partners at Texas State University posit that ontologies and W3C linked data standards and tools may facilitate much earlier impact from the CTE/KDE framework on traceability in the agri-food sector.  This is because these standards were designed to support integrating diverse information, and reason over the results of that integration even when that information is incomplete.

Capability Modeling for Digital Factories (CaMDiF ) project 10 

The objective this project is to enhance the intelligence and effectiveness of various supply chain decisions through providing real-time, dynamic insight into the technological capabilities, capacities, and quality history of manufacturing suppliers. This project resulted in creation of a cloud-based software solution for manufacturing capability modeling and sharing supported by a formal ontology.  This project was conducted in collaboration with the Applied Research Institute (ARI) at the University of Illinois at Urbana- Champaign (UIUC), Indiana Technology and Manufacturing Companies (ITAMCO), and the Innovation Machines.

Smart Manufacturing through a Framework for Knowledge Based System  Diagnosis: project 9

Within the manufacturing industry various techniques are used to diagnose problems throughout all levels of the organization. Often times, this root cause analysis process is ad-hoc with no standard representation for artifacts or terminology used in the analysis. Once a problem is diagnosed and treatments are assessed, the results are discarded or stored locally as text documents or in paper form. When the same or similar problem occurs again in a different factory or with different employees, the whole process has to be repeated without taking advantage of knowledge gained from previous problem(s). Furthermore, when discussing the diagnosis, employees might have miscommunication over terms used in the root cause analysis leading to wasted time and errors. To alleviate these issues, this project is aimed at developing a framework for a knowledge based system of diagnosis. By synthesizing diagnosis methods used in manufacturing and in the medical community, this work proposes a framework which integrates and formalizes root cause analysis that spans multiple organizational levels. The proposed framework aims to leverage recent advances in machine learning for diagnosis in smart manufacturing. In particular, Bayesian Network will be used for developing the probabilistic model.  A diagnosis ontology and a thesaurus are among the main deliverable of this project.

Industrial Ontology Foundry (IOF) project 8

The primary goal of the IOF is to create a suite of open and principles-based ontologies, from which other domain dependent or application ontologies can be derived in a modular fashion, remaining ‘generic’ (i.e., non-proprietary, non-implementation specific) so they can be reused in any number of industrial domains or manufacturing specializations. Dr. Ameri is the chair of the Supply Chain Working Group at IOF.

Other IOF Goals:

• Providing principles and best practices by which quality ontologies can be developed that will support interoperability for industrial domains,
• Instituting a governance mechanism to maintain and promulgate the goals and principles,
• Providing an organizational framework and governance processes that ensure conformance to principles and best practices for development, sharing, maintenance, evolution, and documentation of IOF ontologies.

Text Mining Techniques for Manufacturing Supplier Classification and Clustering project 7  

The web presence of manufacturing suppliers is constantly increasing and so does the volume of textual data available online that pertains to the capabilities of manufacturing suppliers. To process this large volume of data and infer new knowledge about the capabilities of manufacturing suppliers, different text mining techniques such as association rule generation, classification, and clustering can be applied. This research focuses on classification and clustering of manufacturing suppliers based on the textual description of their capabilities available in their online profiles. Different techniques, such as  Naïve Bayes method will be adopted and implemented using R programming language in this project. Casting and CNC machining are used as the examples classes of suppliers in the early stages of the project but more complex classes (such as a class of precision machining suppliers with screw machining capabilities) will be investigated in the future. The performance of the proposed tools will be evaluated experimentally based on the standard metrics such as precision, recall, and F-measure. Also a thesaurus based on SKOS will be used to improve the intelligence of the classifier tool.

Past Projects: 

Measuring Information Content of the Artifacts of Early Design:  project 6

The objective of this NSF-sponsored research is to measure the information content of different artifacts used in early design in a manner that is neutral of the representation and also to empirically study the growth of information, or reduction of uncertainty, in the early stages of engineering design process.   Information of various types is used and produced in design, and each may be expressed in different forms, such as pictorial, text, symbolic, or virtual.  We believe that information is a medium-neutral and form-neutral quantity, i.e., the same assertion, whether expressed in text or sketch, contains the same amount of information and this amount could be measured.  To this end, we will first propose a form-neutral representation based on an ontological approach that could be used to translate any type of design information into a neutral form.  Next, we will develop multiple metrics to measure the information content of artifacts based on this neutral representation and also will develop the necessary automated tools for quantification of the information content of design artifacts.  Previously, we proposed a set of models to measure how decisions, enabled by new information, reduce the design space and thus the uncertainty of the solution.  We will use these models to validate the information metrics by experimentally measuring how information within different artifacts, as measured by the metrics, corresponds to the role of the information in reducing uncertainty.

Collaborative Reference Model Evolution for Smart Manufacturing Systems: project 5

Distributed intelligence is one of the key distinguishing features of the already emerging, next generation manufacturing systems. The new factories of the future are supported by service-oriented, data-driven, and knowledge-based systems. Such factories are operated and controlled by networks of interconnected devices and processes with embedded intelligence. The embedded intelligent agents provide and consume a wide range of services such as process selection, production planning and control, inventory management, quality control, and maintenance planning. A service-oriented manufacturing enterprise creates a network of loosely-coupled processes that can be configured and reconfigured in a timely manner in response to changing needs. This enhances the overall responsiveness, agility, and resilience of the system. Although the idea of Service-oriented Manufacturing (SOM) has been around for more than a decade, there are still multiple infrastructural issues that need to be addressed before the real benefits of SOM can be realized in practice. One of those issues is the development of shared reference models that are used for representation of the services. A reference model, like any other ontology, is a living entity and it evolves over time due to change in the conceptualization of the domain, business strategy evolution, or new user’s needs. The objective of this NIST-sponsored project is to propose a formal process for community-based reference model evolution and identify the computational components required for effective model evolution.

Knowledge Representation in the Cognitive Factory: project 2

Due to increase in the number of product variants in the market and trend toward customized products, manufacturing facilities are increasingly becoming more flexible and reconfigurable in order to be able to respond quickly to changes in work orders. Although Flexible Manufacturing Systems (FMS) and Reconfigurable Manufacturing Systems (RMS) have significantly improved adaptability of manufacturing facilities, they still depend on off-line programming by human programmers. Therefore, the response time of the entire system is prolonged due to limited capabilities and knowledge of human programmers. To fill this gap, manufacturing systems need to be equipped with embedded cognitive capabilities such as self-awareness, sensing, reasoning, and learning. With intelligence and cognitive capabilities embedded in manufacturing systems, they can be reprogrammed and reconfigured autonomously with minimal human interaction. A key component of a Smart Factory is the knowledge model which consists of models of manufacturing processes, models of machine capabilities and models of workpiece.

This research area represents a family of projects that are geared toward developing the components of a Smart Factory with emphasis on knowledge modeling and reuse. Students with strong background in programming (particularly JAVA language) and software development are invited to join this family of projects. Familiarity with CAD/CAM and agent-based technologies is desirable.

Manufacturing Service Description Language (MSDL): project 1

Manufacturing Service Description Language (MSDL) is an OWL-based ontology developed for formal representation of manufacturing services. MSDL was first released in 2006 in PLM Alliance research group at the University of Michigan and further updated and extended in the Infoneer Reseach Group at Texas State University. Currently MSDL is curated at Semantic Computing lab at ASU under supervision of Farhad Ameri.  MSDL is a formal ontology since it is contains explicit semantics coded in a logic-based formalism. OWL-DL, a sub-language of OWL, is selected as the ontology language of MSDL.  OWL is recommended by the World Wide Web Consortium (W3C) as the ontology language of the Semantic Web.  OWL uses XML as the syntax language; hence it has enough portability, flexibility, and extensibility for web-scale applications.  Description Logic (DL) [7] is supported by the Semantic Web meaning that OWL-based ontologies can be shared, parsed, and manipulated through open-source web-based tools and technologies, including multi-agent systems. The original purpose of MSDL was to serve as the ontology language of an agent-based framework for supply chain deployment.