New Delhi. Minus Zero has introduced an end-to-end autopylot system for the first time in India, which enables navigation without lane marking on vehicles coming in dense Indian urban traffic and narrow roads. Testing on some challenging roads in Bengaluru, this vision-based system can handle unique obstacles and their unexpected behaviors such as animals, handcuffs, small 2-wheelers (Yulu), etc.
AI use
To enable such features to unstable traffic, Minus Zero has adopted AI, in which a self-supervised manner has been trained in a self-supervised manner without human label and trained navigation-to-and-found foundational models, instead of traditionally used laws in India inste L1 and L2) With a sudden explosion of systems, the company believes that now more advanced levels such as L2+, L2 ++ and L3 are the right time to shift. Today Indian consumers are ready for more technical-rich features, even if the price is slight increase, and to experience luxury, safety and comfort that comes with it.
Already work in developed countries
While there are already L2+ and L3 systems in developed countries such as FSD of Tesla, Drive Pilot of Mercedes, Super Cruise of GM, GOD’s God’s Eye etc., have been a very challenging task to enable such autopiat systems in unstoppable traffic environment like India and other emerging countries. ADAS products are deployed in some developed countries, 85% of the world’s road accidents come from emerging countries, where such advanced driving support systems are very much needed. In Minus Zero, the company understands this unique pain point and technical change that is necessary to create advanced autopylyt systems for all major emerging countries (starting with India).
First driverless system in India
After showing India’s first completely driverless vehicle for in-campus mobility, the company is now on the leap of the next journey with a full-stack autopylot system for Indian traffic. Autopylot system is on some highly complex landscapes (as is seen in the video) using only cameras and to Navigate on HD Maps without dependence on HD Maps and without relying on HD Maps. Self-supervised takes advantage of the power of learning. This is also the first time End-to-end Foundational models are being tested on Indian roads.
A major result of using end-to-end models is their new scenarios (models have not seen these roads in the first video) and new obstacles (models have not seen the various types of obstacles shown in the video), which make them highly scalable and especially suitable for emerging markets like India and other emerging markets such as unstable and other emerging markets. The company says, “We are already working with some big OEMS to upgrade these solutions and are also interacting with others. While we are still developing and validing the entire system, we aim to get ready for production in the next 2 years with our OEM partners.”
