Instantaneous Driver Safety Assistance and Rear Wheel Maneuvering system based on Computer Vision 2019-01-0095
The paper presents the ideology, concept, design of a prototype based on the adaptive rear-steering of a vehicle to counter under steer especially at corners at very high speeds, to further decrease to a shorter turning radius, improvement in agility at tighter spaces and for example parallel strafe parking and also immediate response to driver steering feedback via the rear axle.
The technology domains utilized in the technology are classified based on the two modules of the system:
Sensor Unit - Marker Identification, in this case chessboard or ArUco marker, Computer Vision and mathematical processing in OpenCV and in Python. Various mathematical concepts are used in the sensing system for accurate detection of axes and calculation of steering angles, and if possible, wheel damping and rebound for suspension stiffening.
Rear Axle Control Unit - Data processed from Sensor Unit is fed to a control system including a microcontroller, rear steering mechanism and also on further development, a rear or four wheel adaptive suspension system based on magnetorheological or fluid based damping.
This approach of a rear wheel system based on computer vision and image processing would decrease the sensing-feedback response time and increase vehicle stability, cornering and agility. This depends on the clocking speed and processing power of the process in use. Various cases and modes consisting of predefined rear steering angles are considered and programmed with respect to the driver input and corresponding to the function required;namely settings involving performance, city commute, parking, comfort etc. More importantly safety due to active rear wheel steering facilitates safe cornering and maneuvering even while at high speeds.
All feedback factors such as steering angle would be as per calculations and geometrical regulations and factors in steering system designs such as turning circles, steering ratios and Ackermann Steering geometry for optimum vehicle dynamics and performance.
Citation: Sengupta, J., "Instantaneous Driver Safety Assistance and Rear Wheel Maneuvering system based on Computer Vision," SAE Technical Paper 2019-01-0095, 2019. Download Citation
Delhi Technological University
SAE Connected and Automated Vehicle Conference Israel