Centre for Automotive Research

Vehicle Intelligent Systems Laboratory

Currently, research in this laboratory focuses on the following areas:

Vision-based Autonomous Vehicle Motion

Target tracking using optical flow

In this work, target tracking is performed using the pyramidal Lucas-Kanade method to estimate the optical flow field from image sequences captured by a camera. The estimated flow field is used to mark an object within the image frame. From the captured images, pixel location of the marked object is used to estimate the desired vehicle trajectory. The process is repeated recursively until the vehicle converges with the tracked object.

Autonomous path planning and obstacle avoidance

Visual data is used to capture the environment in which the vehicle is moving. From the captured image, obstacles are identified and the optimal path to reach the final destination is determined. This is done using a combination of image processing, roadmap generation and optimization techniques.

Active Suspension Design

Suspension design optimisation tool

A tool for optimizing suspension design is currently being developed using the SimulationX/Modelica platform.

Stability and Stabilizability of Vehicles

Modern vehicles are expected to be able to respond quickly to sudden changes in driving conditions, for example, when hitting a wet patch on a dry road or in instances of emergency when the car brakes and swerves hard. Such situations invariably lead to sudden switches in system dynamics. Stability is a major concern in cases such as these. Work is currently being carried out on developing methods and tools, using Maple and MATLAB platforms, for designing controllers that ensure stability of vehicles undergoing abruptly switching dynamics.