Making an all-terrain BAJA vehicle drive itself, as AI Lead for Team Ashvamedha.

role
AI Lead
status
In development
when
Sep 2026 to present · aBAJA 2027
ownership
Team project
abaja.autonomyRobotics & autonomy · Intelligent systems

Who did what. Team Ashvamedha is my university's team for the aBAJA competition under SAEINDIA. I lead the AI side; the vehicle itself is built by the whole team.

Overview

I'm the AI Lead of Team Ashvamedha, my university's team for aBAJA 2027 under SAEINDIA, the autonomous edition of the BAJA all-terrain vehicle challenge. I work on model training and on making the vehicle autonomous: lane keeping (LKA), ODC, multi-lane navigation and point-to-point navigation.

Why it exists

An all-terrain vehicle has to keep its lane, handle multiple lanes and get from one point to another on its own, with every decision made on the vehicle, in real time, and with safety as the first constraint.

My part

  • Train the perception models the vehicle drives on.
  • Work on the autonomy stack for the competition tasks: LKA, ODC, multi-lane navigation and point-to-point navigation.
  • Coordinate how the AI fits with the mechanical and electrical systems the rest of the team builds.

How it works

  1. a

    Train

    Perception models trained for the conditions the vehicle will face.

  2. b

    LKA

    Keep the vehicle centred in its lane.

  3. c

    ODC

    One of the competition's autonomy tasks.

  4. d

    Multi-lane

    Navigate roads with more than one lane.

  5. e

    Point to point

    Drive autonomously from a start point to a goal.

What the car has to do

perceiveplancontrol

I train the models the vehicle relies on and work on making it drive itself. The season is still in progress, so this page lists the tasks rather than results.

  • LKALane keeping: hold the vehicle in its lane.
  • ODCOne of the competition's autonomy tasks.
  • Multi-laneNavigate roads with more than one lane.
  • Point to pointDrive from a start point to a goal on its own.

Built with

  • Model training
  • Computer vision
  • Lane keeping
  • Path planning
  • Autonomous systems

Status & next

Work in progress for the 2027 season. Results will be added once the team publishes them.

Role and scope confirmed by Aaradhya