Development and Field Evaluation of an AI-Driven Virtual Barrier and Intelligent Warning System for Railway Crossing Safety in Uganda

Railway crossings in Uganda,especially the many manned or informal ones along the dilapidated Standard Gauge Railway and metre gauge corridors see a recurring incident of collision or near misses involving vehicles,bodabodas and pedestrians.Manual barriers and staffed crossings cannnot scale to cover the network,and there's currently no low-cost,locally adapted warning system built for Uganda's power and connectivity conditions. This projects develops an A.I driven virtual barrier: low cost sensors(camera,radar/ultra sonic) paired with an onsite A.I model that detects an approaching train and crossing-zone occupancy ,then triggers visual/audible warnings,without depending on continous internet or power connectivity.This project is expected to produce a proof of concept detection and warning system field tested at one pilot crossing,a site characterisation report,a trained detecetion model and A.I research capacity for staff and students .The overall outcome is to prevent accidents at railway crossings.