Fatigue, rash driving? Anna University students' AI model to take the bus wheel

CHENNAI: In a significant innovation aimed at improving the safety of public transport, a group of engineering students from Anna University has developed an AI-powered software application named FleetSense. The system is designed to monitor and enhance the safety of state-run buses in Tamil Nadu.

A John Moses, a third-year Computer Science Engineering student, explained, "FleetSense is an AI-integrated model developed specifically for public transport. If a driver becomes unconscious, distracted, or fatigued during a journey, the system immediately takes control, safely parks the bus, and alerts the control team in real time. It also notifies the vehicle owner or relevant authorities like TNSTC, SETC, or MTC."

The four-member team - John Moses, Kabil Preetham, R Harish, and R Kishor - developed the system in 45 days under the guidance of Assistant Professor A Satheesh. The project was showcased at the Passenger Vehicle Expo 2.0 held at Chennai Trade Centre.

The system uses predictive models to detect potential mechanical issues, such as brake failure, by analysing vibrations and patterns. "Through a machine learning module, the system can assess whether brake components might fail over time, even years in advance. If any anomalies are detected, the system halts the bus before any damage occurs," Moses said.

The driver's condition is monitored using sensors embedded in the steering wheel that measure body pressure to detect signs of fatigue. "If the driver is found to be drowsy or unresponsive, the AI system will take over, keep the bus in the correct lane using onboard cameras, and park it safely without harm to passengers," he added.

K Kabil Preetham, a final-year student, highlighted another innovation: a rover-based system designed to inspect undercarriage components. "Our unit, equipped with deep-imaging cameras, moves beneath the vehicle to detect oil leaks and mechanical faults. The system uses machine learning to process this data and alerts the driver with diagnostic messages and a map to the nearest service centre."

The AI system is currently tested on TATA Prima vehicle models—commonly used in Tamil Nadu's public transport fleets. It monitors fuel efficiency, engine temperature, RPM, and vehicle speed in real time through telemetry. Data is tracked via a central control room, and alerts are also accessible on mobile phones.

The team estimates the installation cost around Rs 1.5 lakh per bus, considering the age of many government buses and the need for high-grade cameras and sensors. "For newer electric buses, the cost will be lower," said Preetham, noting the potential savings in accident-related repair costs and lives.

"We aim to bring this solution to the Transport Department soon. We're also in talks with private operators and are working on filing a patent," he added.

ROAD SENSE

The AI-powered system will track driver behaviour, including vitals, through sensors embedded in the steering wheel

If the driver is unconscious, fatigued or drives rashly, FleetSense will take control of the vehicle and park safely

Vehicle owner (TNSTC, SETC or MTC) and the data control room will be alerted in real time on the status

Primary testing is conducted on TATA Prima models, used widely in TN, and will be developed for other models

Installation per vehicle can cost up to Rs 1.5 lakh, with newer models costing less