Over time, the environment would get more machine friendly with adaptive cruise control, lane assist and automatic braking bringing a shift towards disciplined driving
Bengaluru is tuned to driverless operations; the Namma Metro’s Yellow Line will soon have passengers riding without a human at the controls. The metro succeeds because it operates in controlled, grade-separated corridors with predictable outcomes, whereas driverless cars, must function in open, chaotic ecosystems, where the leap from rail to road is a technological challenge. Bengaluru has some of the worst traffic congestion in the country, with horns blaring, two-wheelers weaving unpredictably and pedestrians crossing with no regard for lanes, posing a daily hazard and a productivity sink.
Given these challenges of technology and terrain, will citizens ever see driverless cars whizzing by on the streets of Bengaluru? With over 7.22 lakh new vehicles added in 2024-25, totalling to 1.23 crore registered vehicles already straining 3,000 km of road network, the city is notorious both for minor scrapes and tragic collisions involving motorbikes, cars, autorickshaws, cyclists and pedestrians. Poor lane discipline, distracted driving, inadequate signage, non-adherence to traffic rules, air pollution from idling vehicles, stress level spikes on long commutes invariably contribute to making Bengaluru’s traffic scenario a virtual hell-hole.
At its core, driverless cars rely on a suite of sensors to detect light, radar, cameras, GPS and advanced computing to perceive the environment, interpret it in real time and make split-second decisions about steering, braking, and speeding. Since driverless cars are designed to eliminate human error, won’t removing the driver actually reduce collisions, ease congestion and transform mobility, making Bengaluru’s roads safer and more driveable? With high-definition maps and machine learning models helping driverless systems predict the behaviour of other road users, shouldn’t driverless vehicles react faster and more consistently than humans, without getting distracted by phones, or getting fatigued after a long day, nor suffering road-rage?
To fit into this spectrum, Bengaluru must provide clear lane markings, predictable traffic patterns, reliable GPS signals and consistent enforcement of rules. Well-structured roads, vehicles moving in orderly, clean, well-painted lanes, with clear traffic signs, orderly infrastructure, zero jaywalking and erratic lane changes must get real. Today’s reality however has vehicles ignoring lanes while scooters, autos and pedestrians intermingle unpredictably. Potholes, faded lines and constructions zones complicate machine vision, with slow-moving vehicles and non-motorised traffic, puzzling machine-learning programs. Where a cow to wander onto a major thoroughfare or an autorickshaw to suddenly swerve without signalling, a driverless car’s sensors would get overwhelmed! To top it all, accident liability, insurance frameworks, data privacy and regulatory standards for autonomous driverless vehicles are still evolving.
Internationally US cities like Austin, are already testing driverless cars with Tesla, Zoox and Waymo Robotaxis navigating Texan streets with a large spinning unit on the car top housing a light detecting and ranging sensor. Austin has already registered an accident when a robotaxi struck a child triggering a Federal Investigation. In India, if not Bengaluru, Chandigarh exhibits adequate urban planning, lower density chaos and better traffic compliance, geometric clarity, a grid system, sector based planning, wide roads, clear intersections, predictable traffic flows, visible lane markings, fewer encroachments and manageable traffic volumes, fit to become driverless. Contrast this with Bengaluru’s old pete areas, IT corridors, colonial-era layouts, gated communities, flyovers stitched onto village roads and ongoing construction, its mixed-mode traffic with SUVs, bikes, buses, cycles, autos and pedestrians, two-wheelers snaking through narrow gaps, autos stopping abruptly, pedestrians crossing wherever opportunity exists. Every stray dog, a stalled bus or a sudden downpour can transform traffic into a congestion hotspot!
And yet, Bengaluru has one advantage: it houses global R&D centres, AI startups, mapping firms, and automotive software engineers. So wouldn’t a hybrid model be more desirable compared to a total technological leap? It could work if certain zones were geofenced and driverless vehicles operated in controlled corridors like tech parks in Whitefield, Electronic City campuses, airport roads, or dedicated business districts where road conditions could be standardised, signage upgraded, and digital mapping made precise. Specially built infrastructure could ensure traffic lights, road sensors and control rooms communicate with autonomous cars and machine readable roads. Over time, the environment would get more machine friendly with adaptive cruise control, lane assist and automatic braking bringing a shift towards disciplined driving. Algorithms need training on Indian road psychology, negotiation, eye contact, honking patterns and informal right of way customs. A vehicle strictly adhering to a red light while surrounding drivers aggressively inch forward will cause friction.
Bengaluru may not become India’s first driverless vehicle city, but the road to driverless Bengaluru will begin when order starts to exist. A hybrid model can impose islands of order within disorder, forcing roads, signals, and drivers to adapt. Change may begin quietly but whether Bengaluru joins that journey will depend not on its coders but on its commuters.
(Priyan R Naik is a columnist and independent journalist living in Bengaluru.)
