End-to-End Foundational AI Models learn to navigate streets of Bengaluru

Minus Zero is glad to unveil an end-2-end autopilot system for the first time in India enabling navigation in dense Indian urban traffic with on-coming vehicles and narrow roads with or without lane markings. Tested on some of the challenging streets of Bangalore, our vision-based system can gracefully handle unique obstacles like animals, push-karts, small 2-wheelers (Yulu), etc. and their unpredictable behaviours. To enable such features in unstructured traffic, Minus Zero has taken an AI-first approach by training end-to-end foundational models that can learn navigation in self-supervised manner from large scale raw data without human labels, instead of traditionally used rule-based systems. 

With the sudden explosion of ADAS (specifically L1 & L2) systems in India over the last few years, we believe it’s the right time to shift to even more advanced levels like L2+, L2++ & L3. Today Indian consumers are ready for more technology-rich features at a marginal increase in the price point and experience the luxury, the safety and the comfort that comes with it. While there are already working L2+ & L3 systems in developed countries like Tesla’s FSD, Mercedes’ Drive Pilot, GM’s Super Cruise, BYD’s God’s Eye, etc., enabling such autopilot systems in an unstructured traffic environment like India & other emerging countries has been a very challenging task. While ~90% of the ADAS products are deployed in few developed countries, >85% of road accidents of the world come from emerging countries, with much more need for such advanced driving assistance systems. At Minus Zero, we understand this unique pain point and technology shift required for building advanced autopilot systems for all the major emerging countries (starting with India). Having shown India’s first fully driverless vehicle for in-campus mobility, we now embark on our next leap of journey with a full-stack autopilot system for Indian traffic. 

Our autopilot system leverages power of bespoke foundational models and self supervised learning to navigate some highly complicated scenarios (as seen in the video) using only  cameras, and without reliance on HD Maps. This is also the first time when end-to-end foundational models are being tested on Indian roads. A great outcome of using end-to-end models is their generalisation capabilities to new scenarios (the model has never seen these roads in the video before) and new obstacles (model has not seen these different types of obstacles shown in the video), which makes them highly scalable and specifically suited for unstructured and dynamic traffic environments like India and other emerging markets. While we have showcased multiple scenarios in this video, there are more that we are still working on like overtaking, unprotected turns, roundabout navigation, etc. It’s important to note that our autopilot system comes under the ‘Hands OFF & Eyes ON’ category where it’s mandatory to have a safety driver onboard.

We are already working with a couple of large OEMs in their journey to upgrade them to these solutions and are in talks with others for the same. While we are still developing & validating the entire system, we aim to be ready for production in the next 2 years alongside our OEM partners.

So what’s different that we are developing from the technology point of view?

• Big Technology Shift - From supervised rule-based algorithms which require heavy manual input to leveraging end-to-end Foundational AI models that can learn from large scale raw data in a self-supervised manner.

• Minimal Hardware - From complicated & expensive sensor-setup including multiple LiDARs to vision-first stack that is easier & more economical to put in production.

• Mapless Autonomy - Removing dependency on building highly precise HD maps that are extremely expensive to build & maintain and practically non-existent in India & other emerging countries currently.

• Scalable from Day 1 - Everyday our system is constantly improving to be a better driver with more data, and will enable us to rapidly scale to new environments, geographies and vehicle form factors without re-engineering the system.

• Full Stack AI Platform - To enable this autopilot system, we have indigenously built full-stack AI solutions including leveraging generative AI and VLMs for understanding data, creating photo-realistic real-world simulations to test & validate the system, and bespoke on-board software to deploy the system safely.

After LLMs and AI agents for the digital world, next-gen foundational models for the physical world like autonomous driving, robotics, etc. are gaining momentum globally. With India taking a strong stance in creating a foothold in the AI race, this area is an opportunity where we can build in India for the world. And Minus Zero is proud to be at the forefront of deep tech innovation in India since inception, spearheaded by a supremely talented team, rapid execution, and a commitment to staying ahead in technological advancements. 

For more information, contact us at team@minuszero.ai.