“Leveraging the true power of autonomy tech and building the product with a lasting impact is how we define success at Minus Zero.”Gagandeep Reehal
For this week’s startup feature, Analytics India Magazine spoke to Gagandeep Reehal, the CEO and CTO at Minus Zero to understand how the Jalandhar-based startup is leveraging emerging technologies to create fully autonomous driving capabilities in electric cars.
Founded in 2020 by Gagandeep Reehal and Gursimran Kalra, Minus Zero is India’s first start-up building affordable fully self-driving electric cars in the country. The company was founded amidst the pandemic and is headquartered in Jalandhar with an R&D unit in Patiala currently.
The founders claimed the company is capable of Level 5 autonomy with highly energy-efficient electric vehicle design and proprietary nature-inspired AI that is less dependent on extensive data & costly sensor suite (like LIDAR’s, etc.).
Reehal said, “Our fully Autonomous Electric Cars will be able to drive autonomously in highly unstructured and disorganised traffic scenarios like India’s.”
Currently, Minus Zero is planning to start pre-orders of the fully autonomous electric car in late 2023. The self-driving car is capable of Level 5 autonomy in the toughest Indian traffic scenarios and is coupled with an energy-efficient electric vehicle design that could cover around 180 km on a 45 minutes charge with a total capacity of 500 km on a single charge.
Reehal added, “Ranged between INR 16-25 lakhs, the vehicle will cater to the needs of Generation Z, with a hybrid sports design and high performance at a competitively affordable price. There will be a lot of features which we will reveal over time.”
What’s The Differentiator
“Many industry prototypes make heavy use of LiDARS and secondary sensors, while our Self Driving Architecture will drive the car similar to how a human does using only eyes, camera in the car’s case. This will make the travel experience more rider-centric, eliminating the pressure on a driver, and enabling safer travel,” said Reehal.
He also mentioned some of the competitive advantages
- The company’s perception algorithms do not depend on lane markings for road detection, making it possible to drive autonomously even on roads in bad condition.
- The systems use 10x lesser GFLOPS per frame achieving high-speed processing, even on edge hardware, making it computationally efficient without compromising on accuracy.
- Zero dependencies on expensive sensor suites like long-range LiDARS, even at night.
- Nature-inspired vision AI uses predictive behavioural-based path planning to ensure robust autonomy in even unseen environments.
Use of AI @ Minus Zero
Minus Zero uses a hybrid neural network, taking into account the cognitive bias generated by serial-progressions in an intuitive way, to finetune the prediction. The company’s proprietary approach solves one of the computer vision’s major problems, occlusion. This also decreases the need for more training on a large dataset, thus making it robust even on completely random traffic scenarios.
Core Tech Stack
The tech stack at Minus Zero includes various machine learning techniques, open-source Python-based frameworks. The mechanical design & control of the vehicle are being developed in CAD environments for dynamic testing, design and simulations.
At present, Minus Zero includes a team of 25 employees. While hiring, the company looks for dedicated and passionate individuals with a research bent of mind, who are willing to explore new technology with a strong practical base.
“Our whole focus and energy is devoted to bringing the product into the market as early as possible. As mentioned earlier, we will be launching a market-ready version of our product by late 2023 and file for a regulatory testing post that. We will launch a set number of our initial model in some specified cities and monitor it closely. Along with that, our R&D team will continue to make improvements & bring updates to our self-driving software. We will begin mass-scale production after a successful launch in the specified cities and thereafter, ship out the products throughout India. We hope to achieve our target sales of 5,000 units by late 2025,” said Reehal.
Subscribe to our NewsletterGet the latest updates and relevant offers by sharing your email.
A Technical Journalist who loves writing about Machine Learning and Artificial Intelligence. A lover of music, writing and learning something out of the box.