Formal Verification and Validation of Artificially Intelligent Systems.

Digital Twin framework for space robotic applications, in collaboration with European Space Agency.Formal Verification framework for an autonomous Mars rover with AI/ML components.

Project portfolio:• Collision avoidance with stochastic obstacles in dynamic and uncertain environments for L4 Mobile Autonomous Systems : Spline based Motion Planning and tracking using Model Predictive Control, Behavior modeling and situational assessment. • Automated vehicle lateral and longitudinal control with successful demonstration on an Autonomous Guided Vehicle (AGV) in an industrial warehouse setting.• Hybrid derivative functions that closely combine expert knowledge of the system captured by ordinary differential equations, with data-driven feedforward neural networks on slider-crank mechanism.• Inverse dynamic Nonlinear model of magnetic-spring assisted 5-bar SCARA robot to improve control performance during pick and place operation. Identification of unknown force in physical dynamic model by Deep Feedforward Neural Network to further improve control performance.• Created sequence modeling methodology using Long Short-Term Memory (LSTM) networks for Automated parameter adaptation in an iterative hot embossing process using machine generated and manually logged operator data.• Energy management strategy and control of drivetrains equipped with Electric Variable Transmission (EVT) for industrial machines and off-highway vehicle applications.• Object-oriented toolbox development for optimal topology, sizing and control of electric and hybrid powertrains with multi-energy sources/storages for automotive applications using Dynamic Programming and ECMS.

Optimal Sizing of Waste Heat Recovery System (WHR) for Dynamic Engine Conditions.Optimal sizing of the Waste Heat Recovery system is an important aspect in terms of maximizing its power output and efficiency, while also minimizing the system weight, losses and fuel consumption of a heavy-duty vehicle. Performance evaluation on steady state operating points by means of driving cycle reduction tends to overestimate the fuel gain induced by the WHR system when compared to the dynamic predictions. In this study, a methodology for optimal sizing and control of waste heat recovery system is presented over hot-start World Harmonized Transient Cycle. The criteria for WHR system to be optimally sized is maximization of the output power of WHR system or fuel gains from the engine. A scalable WHR system model is developed based on sensitivity analysis of different components with most significant impact on power output of the system due to sizing. The components scaled in this study are EGR and exhaust gas evaporators, and expander.The key challenge in developing a methodology is the coupling between design and control parameters. An alternating optimization architecture is developed for optimal design and control of WHR system for transient driving conditions while satisfying safe WHR operation constraints.The optimal component sizes of WHR system are found to be different for different driving conditions. A switching model predictive control (MPC) strategy is implemented on optimally sized WHR system, which offers an additional fuel reduction as compared to the benchmark system over hot-start WHTC.

Optimal Sizing of Waste Heat Recovery System for HD truck: Steady State Analysis• Developed methodology to obtain an optimum size of WHR system for steady-state driving conditions.• Developed scalable model of WHR system.• Developed size independent low-level controller.• Applied nonlinear optimization techniques to minimize fuel consumption.

• Base calibration of Suzuki engine on dynamometer using model-based design.• Engine ECU calibration on Suzuki vehicle for air/fuel system monitoring and meeting OBD2 emission regulations.• Complete Bi-fuel system set-up (CNG + Gasoline), and ECU calibration on engine bench and proto vehicle. Performance benchmarking with factory system.• Endurance testing and fault diagnosis of prototype CNG fuel injectors on gasoline engine.• Cold and Hot weather emission testing of Suzuki vehicle on chassis dynamometer: Set-up, analysis, and calibration.• Awarded for most technical reports in R&D Department.