Campaign Story

VAIDYUTH started with something the team saw every single day: their campus groundskeepers, maintenance workers, and cleaning staff hauling heavy equipment across the grounds by hand — strenuous, slow, and physically punishing, with no solution in sight. So a group of undergraduate engineers at Sahrdaya College of Engineering and Technology (SCET), Thrissur, decided to build one. Not a slide deck. An actual autonomous vehicle.

What they’ve built is a four-wheeled electric vehicle that navigates a mapped campus with zero driver — switchable between carrying cargo and carrying passengers, with a wheelchair bay planned at the rear, making it the first autonomous, accessible, campus-built vehicle of its kind in Kerala. Under the hood it runs a full production-grade robotics stack: ROS 2 Jazzy on Ubuntu 24.04, SLAM mapping via slam_toolbox, Nav2 for path planning, GPS + IMU sensor fusion, OpenCV for lane and obstacle detection, and a custom serial protocol bridging the ROS 2 brain to an ESP32 motor controller.

And it works. The team has already assembled the vehicle, brought the entire ROS 2 node architecture online, and successfully completed indoor SLAM mapping and night-time outdoor mapping — all documented with data logs and video. They did this on ₹32,550 of their own money, before asking anyone for a rupee.

There’s exactly one thing standing between VAIDYUTH and full daylight campus autonomy, and it’s hardware, not software. Their indoor LiDAR (YDLidar G2) is blinded by direct sunlight — solar infrared overwhelms its photodetector, so it maps flawlessly at night and returns garbage in daytime. It’s a known physical limitation of that sensor class; no software fix exists. The second constraint: a Raspberry Pi 4B can’t run SLAM, Nav2, and computer vision simultaneously without choking. The fix for both is concrete and cheap — a daylight-rated outdoor LiDAR (RPLIDAR A3/S2) and a GPU compute module (NVIDIA Jetson). That’s the entire ask.

The bigger picture is where it gets serious. The autonomy stack VAIDYUTH proves out becomes VaidyuthOS Board — a plug-and-play autonomous-vehicle controller card that packages the whole SLAM/Nav2/sensor-fusion/CV system into one pre-configured unit that drops into any EV chassis with a standard CAN/UART interface. Today, every AV builder has to assemble a ROS 2 stack from scratch over months. VaidyuthOS Board hands them a ready-to-deploy brain. The target market is real and large: 5.1 million+ electric 2W/3W on Indian roads, a ₹22,000-crore last-mile logistics sector, campus and hospital fleets, and 30+ addressable EV three-wheeler OEMs (Bajaj RE, Mahindra Treo, Piaggio Ape EV, Euler Motors). The global campus AV market alone is projected past USD 4.2 billion by 2030.

The wedge is price. Global autonomy rivals like Navya and EasyMile sell at ₹50 lakh+ per unit. VaidyuthOS Board targets ₹45,000–₹80,000 — built on an open-source ROS 2 foundation that cuts IP risk, hardware cost, and development time. Revenue stacks across hardware sales, on-site installation, annual SLA, white-label licensing, and multi-vehicle campus fleet contracts. This ₹70K–₹1.1L raise is the bridge that takes a proven night-time prototype to a daylight-operational vehicle — and a daylight-operational vehicle is what unlocks the next tier of capital (KSUM, DST-NIDHI PRAYAS, AIM-ARISE — ₹5 lakh to ₹50 lakh in grants the team is already lined up to pursue).

Students. A real campus. A real problem. A working robot. The smallest ask on the platform with some of the biggest leverage behind it.