Dr. Amir Kashani | Director, AI & Digital Product | Stanley Black & Decker
Dr. Amir Kashani has over 12 years of experience in applied machine learning in the areas of data-driven modeling for condition monitoring, predictive maintenance, risk-based decision making, and automation across various industries. He has 32+ peer-review publications, 3 US patents pending on AI systems, he has been awarded by multiple federal grants in Applied Machine Learning and IoT (NSF, DOE, DOD) and is also an AWS certified practitioner. At SBD, he is leading the technical development of Stanley Industrial Services for Remote Monitoring and Intelligent Diagnostics and Systems Health Management for the Industrial segment including Automotive, Advanced Tools, Infrastructure and Inventory Management. He has built and led cross-functional teams that have delivered and commercialized breakthrough multidisciplinary AI-driven B2B and B2C products.
Sundeep Kumar | Head of Products & Big Data Services | Sigmoid
Sundeep Kumar has over a decade of experience in the data & analytics industry and has worked with global IT service giants such Mu Sigma, Cognizant Technology, and Infosys. Sundeep has dedicated his career to using data analytics to improve supply chain management through process automation & excellence. At Sigmoid, he has led multiple MLOps and AI related projects in the manufacturing sector to help drive better business strategy.
Achieving Federated Predictive Maintenance - a Data Engineering roadmap
Recent research has shown that automobile vehicle model rollback is a multi billion cost every year. IoT and AI-driven analytics have transformed how companies approach equipment creation, maintenance and production line management. With sensors capable of collecting real-time data, predictive maintenance enables organizations to effectively anticipate and plan for events such as faults, part replacements or equipment breakdown.
Stanley Black & Decker, has thousands of machines installed across Automobile companies plants to automate processes such as welding and bolting. These machines produce IoT data from the sensors which were monitored by QA Analysts, Plant Managers, and the Procurement team on a central dashboard with the objective of predicting fault occurrences, machine failures, and other downtimes.
Problems that needed to be addressed:
Real time flagging of issues with the help of near real time dashboards
Ability to slice and dice the data at various levels to understand fault pattern
Programmatically identify faulty parts and prevent them from being added to the finished vehicle
Sigmoid created a solution for both Cloud and On Premise for SBD that enables:
Central monitoring and diagnosis of machines while keeping the cost of integrating more machines into the system low and consistent
Near real-time updates of dashboards
Better forecast the probability of downtime and malfunctioning of machines