CInDI Computational Intelligence & Design Informatics Laboratory
Design Engineering Research at Wayne State University
The Industrial Revolution
In the heart of downtown Detroit, Michigan, our team of graduate students and post-doctorates are forging the design engineering research.
Wayne State Department of Industrial and Systems Engineering professor Kyoung-Yun Joseph Kim leads our ever-changing team of students and post-docs, but the mission stays the same: To empower engineers and designers with data.
Linear Vs. Relational
Our first hurdle has been to see the forest through the trees. In traditional product design, models are developed in a sequential, linear manner, based on a product’s life cycle. However, they are not easily transposed into systems with conflicting frameworks. Moreover, it's difficult to remove a portion from the sequence without losing its context.
To address these limitations, our team applies mereotopology in the design of our systems. Mereotopology, a mathematical theory for mapping the relationships between parts to a whole, provides a framework for us to parse and employ the complex relationships between parts of a design process and the finished whole of a product. Our designs focus not on the sequences of parts, but on the relationships between them. This paradigm shift allows us to create systems that are modular, interoperable,and machine-interpretable.
Semantics Based Reasoning
While our mathematical approach ensures a greater level of interoperability, our semantic methodologies improve our design systems’ accessibility and performance.
Conventional, knowledge-based approaches are based on the accrual of increasing levels of expert knowledge. At CInDI, we are implementing the more adaptable system of logic that is based on semantic reasoning.
Our semantics-based system addresses the reported limitations of Knowledge Formalization and Representation (KFR)—AI systems that depend entirely on the kind of extensive knowledge base only a machine can contain. Limitations of KFR systems include:
- Challenges for individuals who aren’t logicians, either by nature or training
- Inefficient use of code
- Diminished interoperability with systems lacking the same levels of knowledge input.
Semantics-based systems rely not on encyclopedic knowledge, but on a few key ontological definitions combined with a capacity for contextual reasoning. Our system’s resonance with human learning promises superior levels of flexibility and adaptability than knowledge-based systems. This improvement allows us to create simpler and more efficient models that is intuitively grasped by users of the system, and easily incorporated into other systems.
Real World Applications
The research we do at CInDI pays off. Right now, we are creating a breakthrough, data-driven design framework for manufacturers who use metallic assemblies. System Integrators and OEMs currently have to rely on suppliers and testing companies to physically test each new material they are considering.
Training the Next Generation
While we are excited about our current projects, we are keeping the future of design and manufacturing envirionment our top priority. To that end, we are developing a Digital/SMART Demonstration Center (WSU D/SDC) for the high bay of the manufacturing engineering building at Wayne State University’s College of Engineering. This demonstration center will provide both students and researchers the opportunity to interact with real-time manufacturing, inspection, design/engineering data, and fully-sensored, advanced manufacturing processes.
Our team is currently establishing partnerships with various industries to advance the development of this state-of-the art, industry-supported manufacturing environment. This center will provide a one-of-a-kind opportunity to educate, demonstrate, and research the Internet of Things (IOT), Digital Manufacturing, SMART manufacturing, and real-time manufacturing analytics.