Caterpillar receives data from about one million connected assets and applies advanced analytics to this data to provide prognostic alerts to many customers of impending maintenance needs. These alerts are provided to one or more tiers of expert advisors who examine them together with other data to filter the most significant alerts, which the customer may or may not act on them. Better data on which alerts are send to the customers and acted upon is critical to improve the accuracy of analytics and to understand the value they provide. Unfortunately, this closed loop feedback is often not available.
However, Caterpillar does have access to parts invoice and work order data and can usually map these to particular assets and maintenance tasks. The contents of these work orders and their timing relative to alerts may be mined to provide a better understanding of when an alert was acted upon or, if an alert was ignored, the negative consequences of doing so.
Student team will develop a data model that accurately predicts customer response to maintenance alerts on representative equipment, determine the value of particular alerts (e.g., repair dollars saved, downtime avoided, etc.), and finally to generalize/characterize the business conditions where alerts deliver the highest savings. Results will be linked to IoT prognostic analytics models to provide more accurate ground truth for training and tuning these models.
This project will improve ongoing accuracy improvement of IoT prognostic analytics and, by quantifying the impacts of these analytics, will enable the development of innovative new business models.
User Interface Design (1 student)
Specific Skills: UI/UX Design, usability studies, graphic design experience/interest desirable
Likely Majors: SI (MS), A&D, CS
General Programming (2-3 students)
Specific Skills: General Programming skills, good software engineering practice and design. Must have completed EECS 281 prior to W22.
Likely Majors: CS
Analytics / Machine Learning (2-3 students)
Specific Skills: Experienced with machine learning and classification. Please highlight your specific experience in your personal statement.
Likely Majors: CS, DATA, IOE, SI (MS)
Business Development (1-2 students)
Specific Skills: Customer Needs analysis, branding and marketing, business metrics, focus group testing (These students must have basic coding skills)
Likely Majors: BBA with CS Minor, CS with BBA minor, IOE
Manav Das is a Data Scientist in Cat Digital where he utilizes telematics and transactional data to provide solutions in areas ranging from component life extension to mine-site productivity. Before joining Cat Digital Manav was a Team Lead in the Virtual Product Development (VPD) organization within Caterpillar where he developed and deployed physics-based simulation tools to predict machine performance and durability. Manav has a BS in ME from the Indian Institute of Technology, Kharagpur and a MS and PhD also in ME from University of Arizona. He is currently enrolled in the MS in Analytics program at Georgia Tech.
Steve Buster joined Caterpillar in 1997 as a Program Analyst working on dealer ERP integration projects. Since then he has held numerous roles within application design and development, infrastructure design, and platform architecture. His primary business focus has been in the asset services and operations domain with a focus on enabling our dealers and customers optimize machine health, utilization, and productivity. Steve has a passion for all things technical and enjoys finding new and innovative approaches to apply cutting-edge technologies to real-world problems. Past projects have allowed him to apply technologies such as IoT (Internet of Things), Big Data, and Cloud Computing to a broad spectrum industry solutions.
Executive Sponsor Mentor
Jeff Krupp joined Caterpillar in 2019 as the Director of Applications for Cat Digital. Within his role at Cat Digital, he works with his Digital peers and industry partners to execute Caterpillar’s cohesive digital strategy overseeing the development and user experience for Cat Digital’s broad portfolio of applications that support Caterpillar’s dealers and customers. Prior to Caterpillar, Jeff has served as the Chief Information Office of Agent & Broker Technology, SVP of Technology for OEM products and services at FordDirect and has a 10year career with Ford Motor Company. Jeff holds a bachelor’s degree in Computer Engineering and an MBA from the University of Michigan.
Executive Sponsor Mentor
Dan Reaume is the Chief Analytics Director at Caterpillar. In this role, Dan leads a world-class organization of data scientists, mathematicians, developers, and statisticians designing and developing advanced predictive analytics and optimization algorithms to drive greater competitive advantage value for Caterpillar, its dealers, and its customers. Dan provides strategic and tactical leadership to set technical priorities, improve modeling practices, maximize value generation, and deepen expertise in key practice areas. Prior to joining Caterpillar, Dan led Revenue Analytics’ analytics organization as Vice-President of Operations Research; led pricing and customer experience analytics efforts for Dow Chemical; was the founding director of Advanced Analytics for Dow Corning; and served as technical fellow and leader of the Senior Leadership Technical Council for General Motors. Dan holds a bachelor’s degree in Mathematics and Computer Science from the University of Windsor, Master’s degree in Management of Technology from the University of Waterloo and a PHD/ MS from the University of Michigan in Industrial and Operations Engineering, intelligent Transportation Systems. Industrial Engineering at Wayne State University in Detroit, MI, and his MS in Industrial and Operations Engineering at the University of Michigan in Ann Arbor, MI.
Vineet R. Kamat
Professor, John L. Tishman CM Faculty Scholar
Department of Civil and Environmental Engineering
Vineet Kamat is a Professor of Civil and Environmental Engineering at the University of Michigan. His group researches methods to enable effective human-robot work collaboration in the construction, operation, and maintenance of civil infrastructure and the built environment. Our research has developed several licensable technologies that support automation and robotics, including modeling techniques that help on-site construction robots with autonomous decision making.
Course Substitutions: Honors, ChE Elective, CS MDE/Capstone, CE MDE, Data Science Capstone, EE MDE, IOE Senior Design, IOE Grad Cognate, SI Elective, SI Grad Cognate
Citizenship Requirements: This project is open to all students on campus. Students applying will be subject to Caterpillar compliance screening in order to participate on the team.
IP/NDA: Students will sign standard University of Michigan IP/NDA documents.
In Person/Remote Participation Options: Most of the work will take place on campus in Ann Arbor. If it is safe to travel, the students will have the opportunity to visit the Cat Digital facilities in Chicago and Peoria, Il, as well as local dealers. MDP will provide transportation.
Students who are approved to attend classes remotely for Winter (and Fall if necessary) 2022 may participate on this project.
Internship/Summer Opportunity: Students will be guaranteed an interview for a 2022 internship. The interviews will take place in Jan/Feb 2022.