AI in Ground Transportation

Bengaluru-based D2C grocery platform GreenCart replaced a static 2022-vintage dispatch system with an AI route-optimization engine, cutting cost per delivery 34% (₹42 → ₹28) and lifting metro-peak SLA adherence from 82% to 96% across 18 cities in 20 weeks.

Uber Eats overhauled its homefeed recommendation model, adding transformer-based user behaviour sequences to its DCNv2 model and near-real-time features from its UserContext platform that cut data lag from days to seconds. It is now moving to a transformer-centred listwise generative recommender, supported by a PyTorch migration, offline TensorRT conversion and CPU/GPU disaggregation.

VersaFleet deployed its AI-powered dynamic route optimization solution to help logistics companies reduce delivery times, lower fuel costs, and improve fleet utilization through real-time route planning and traffic-aware algorithms.

JR East said it will start a trial in April 2026 of an AI system that analyses camera images of pantographs on Yamanote Line trains near Shimbashi, Ebisu, Mejiro and Uguisudani stations in real time and notifies the control room if it detects damage. Combined with drone inspections of overhead wires, it aims to cut the time to resume service after equipment problems by 30%.

Moia, Volkswagen's Hamburg ride-pooling service, is testing emotion3D's Cabin Eye interior camera software on its self-driving ID. Buzz AD shuttles. The AI checks doors, floor obstructions, lighting, passenger count and seatbelt use, then signals the vehicle to depart or hands over to fleet control; safety drivers stay on board in testing.

Kiwi.com uses AI and machine learning algorithms to optimize flight search results, predict price changes, and recommend optimal travel itineraries for customers booking flights worldwide. The AI system processes millions of search queries daily to deliver personalized travel recommendations and detect booking anomalies.