VIRAT
2019 - Autonomous Ground Robot
2nd Overall Globally - IGVC 2019
2nd Position - Interoperability Challenge
5th Position - Auto Nav Challenge
6th Position - Design Challenge
PERCEPTION & SENSING
-
Lidar
-
GPS
-
IMU
-
Cameras
-
Encoders
Computation & Actuation
-
Intel NUC i7
-
Motor Driver
-
PMDC Motor
Mechanical Design Features
-
Fully Autonomous
-
Manual Tele OP
SOFTWARE
The software code is written using ROS Framework to enable end-to-end GPS way point navigation and obstacle manoeuvring. Position estimation of the vehicle is reinforced by fusing sensor feedback more efficiently through EKF. Novel, scene-independent and robust path planning algorithms for lane, pothole and flag detection have been developed and integrated.
ELECTRICAL
The electrical module incorporates precise motor position feedback mechanisms and a state-of-the-art battery management system. The lower controller framework provides for efficient processing and debugging of five sensor inputs: (i) Light Detection And Ranging (LIDAR) sensor for 180 degree obstruction detection (ii) Inertial Measurement Unit for route planning in GPS (iii) Cameras for lane detection
MECHANICAL
The bot is from a highly modular aluminium chassis with robust motor and gear coupling. The frame has been designed with a focus on efficient space and low centre of gravity for stability. The current physical design is formulated so as to meet the prequalification requirements of the Intelligent Ground Vehicle Competition.