- Actively working as a Senior Controls ADAS Engineer at Dura Automotive Systems responsible for algorithm and feature development based on customer requirements.
- Technically well versed with several control techniques such as PID, LQR, MPC for longitudinal and lateral vehicle control
- Worked with vehicle dynamics models in PreScan, CarSim, MATLAB/Simulink actively developing vehicle throttle/brake/steer.
- Worked actively within the Unity Simulator with ROS packages. Worked with the gazebo environment and in writing ROS nodes for the drive by wire systems on autonomous cars.
- Implemented a speed and steering controller based on PID gains, stanley controller, MPC based control.
- Fine tuning the PID gain values was implemented using Ziegler Nichols tuning methods, twiddle and several manual iterations based on desired output.
- Currently working with MPC based control for lane changes, lateral steering control. Moreover also working with a deep learning end to end learning based, reinforcement learning based control approaches simultaneously.
Here is an example of initial gains to the PID controller keeping derivative and integral gains to zero. We can see oscillations in lateral vehicle behavior.
Here is an example with initial tuning of the PID controller understanding the effects of Kp, Ki and Kd gains on steady state error, overshoot and settling time properties of the output. We can see an improvement in the lateral vehicle behavior over the same turn now.