India’s First Vision-Only Autonomous Car Was Built for Chaos

On a frenzied Bengaluru morning, a cement mixer veered across three unmarked lanes. A family on a scooter threaded diagonally through the traffic. A pedestrian signalled an auto-rickshaw with a sideways glance and a certain moral authority. Amid this chaos, a white autonomous prototype glided forward. This was zPod, an autonomous vehicle developed by Bengaluru-based startup Minus Zero, and it is unlike anything else in the global self-driving ecosystem. There’s no LiDAR (Light Detection and Ranging) turret scanning the horizon, just a ring of high-resolution cameras, and an artificial intelligence system trained to reason through uncertainty. “For AI to work for humans,” co-founder and chief executive officer Gagandeep Reehal said in a recent interview, “It’s got to think like a human.”

Unlike Tesla’s vision-plus-radar fusion or Waymo’s high-definition LiDAR stacks designed for the orderly highways of California, zPod is made for the utter unpredictability of Indian streets. It doesn’t wait for clean lane markings. It interprets gestures, infers intent, and adapts mid-action. In one test on Bengaluru’s Outer Ring Road, a stray cow paused in the middle of rush-hour traffic. zPod slowed, recalibrated, waited, and moved on—no sudden braking, no algorithmic paralysis.

Minus Zero is a homegrown startup founded in 2021 by Gagandeep Reehal and Gursimran Kalra, whose prior experience in autonomous vehicle (AV) systems abroad gave them a sharp view of what wouldn’t work in India. Backed by Chiratae Ventures and JITO (Jain International Trade Organisation) Angel Network, and operating on a seed round of USD 1.7 million, Minus Zero has positioned itself as one of the few deep-tech ventures in India with a credible shot at shaping the global conversation on autonomous driving—by solving for the hardest roads first.

The company’s approach to autonomy is radical because it is grounded. Minus Zero’s foundational model was initially trained on millions of unlabelled video frames—dashcam footage, traffic feeds, even blurry, rain-streaked windshield recordings—and refined using human expert driving behaviour. Unlike legacy AV technology stacks built around separate modules for detection, prediction, and planning, zPod runs on a single end-to-end model. There is no pause between seeing and deciding. The car thinks as a whole. The result is a system that doesn’t just recognise obstacles; it anticipates how they might behave. While Tesla’s Full Self Driving (FSD) still operates under close driver supervision, and Waymo’s geofenced vehicles depend on intricate three-dimensional maps, zPod moves without either. It handles Yulu bikes, pushcarts, unpainted intersections, jaywalkers, and roadside goats, all with nothing more than vision and judgement.

But despite its technological leap, Minus Zero isn’t aiming to launch a consumer-facing autonomous car, at least not yet. India’s current legal framework under the Motor Vehicles Act prohibits fully driverless vehicles on public roads. In December 2023, Union Minister for Road Transport and Highways Nitin Gadkari reaffirmed that autonomous cars would not be allowed “in the foreseeable future”. This prompted a strategic pivot. Minus Zero is now positioning itself as a software platform provider, offering its True Vision Autonomy (TVA) stack to Indian original equipment manufacturers (OEMs) looking to build advanced driver-assistance systems (ADAS) in the Level 2+ and Level 3 autonomy range.

Already, the company has signed a memorandum of understanding with commercial vehicle giant Ashok Leyland and is in talks with Tata Motors, working to embed TVA into trucks, buses, and eventually passenger electric vehicles (EVs). ADAS offerings in India—like those in the MG Astor and the Mahindra XUV700—use imported radar-camera stacks built for well-marked roads. Owners and testers report frequent lane-keep disengagements or aggressive corrections on unstructured or faded road markings, confirming that these systems struggle in real-world Indian conditions.

Reehal has said that if Tesla is like iOS—an integrated, premium, tightly controlled ecosystem—then Minus Zero wants to be Android: adaptable, open, and ready for the fragmented, chaotic realities of emerging markets. India is the perfect crucible for such a product. The system is designed as “Hands Off, Eyes On”, which means a human driver remains alert, but the car manages key driving functions such as steering, braking, navigating intersections, and reacting to unpredictable agents. Unlike conventional ADAS, which struggle in unstructured environments and depend heavily on HD maps and geofencing, TVA’s reasoning-first architecture allows it to generalise across cities and contexts. What it learns in Bengaluru, it can apply in Jaipur, Jakarta, or Johannesburg. As Reehal put it in a post on X: “Driving isn’t about perfectly matching one ‘ground truth’ path — it’s about choosing from many valid, safe, and socially acceptable options in a complex world.”

Minus Zero wants to go beyond Indian roads to tap global markets. With over 85 per cent of the world’s road accidents occurring in emerging economies, the company sees immense potential for TVA in Southeast Asia, Latin America, Africa, and the Gulf. And the bet is not just on autonomy—it is on accessibility. By using cameras instead of LiDAR and radar, TVA cuts hardware costs dramatically, making AV solutions viable in price-sensitive markets. While US  and Chinese players often depend on massive capital outlays, simulation-heavy R&D, and mapping infrastructure that doesn’t exist in most of the world, TVA is trying to drive where others cannot even test.

The startup is not alone in trying to crack the Indian autonomy puzzle. In Bhopal, Swaayatt Robots—founded by Sanjeev Sharma in 2015—is developing deep reinforcement learning-based driving systems for unstructured environments. The company has demonstrated obstacle avoidance, bidirectional traffic negotiation, and map-free navigation. In Bengaluru, Flux Auto works on retrofitting trucks with highway-ready ADAS features such as adaptive cruise control and lane-keeping assist, while Ati Motors deploys fully autonomous electric cargo vehicles in closed industrial environments like warehouses and factories.

What sets Minus Zero apart isn’t just the tech but the refusal to wait for the world to become orderly. In a field crowded with expensive, overengineered dreams, zPod is a quiet, camera-fed assertion: if you can drive in Bengaluru, you can drive anywhere.

ADAS featuresAutonomous CarHD mapsLiDARnavigationTeslatrafficzPod