Amalgamation Of Internet Of Moving Things (IOMT) And Artificial Intelligence (AI)
With computer vision gaining momentum in the driverless car segment, majority of automobile giants have invested in their R&D. What is the scenario for self- drive cars in India?
Computer vision is only beginning to gain traction in India, especially in the self-drive and car rental category. In order to maintain the economic viability of self-drive cars, reducing accidents and damage occurrences is the key. This is where computer vision comes handy and zoomcar is running large scale experiments for its implementation. The applications in this area include collision detection, tracking driver behaviour, speed correction as per speed limits, assisted actions in case of bad weather and sending alerts in case of major accidents. All Zoomcar’s are IOT enabled and most have AI powered dashboard cameras integrated in them. These help us gauge the optimum health of the vehicle and delve deeper in the segment to confront various challenges in the self-drive and driver safety category.
What drove the industry to start evaluating an AI-based solution? What was the current driving patterns and gravity of the problem?
Necessity is the mother of all inventions. For us, the necessity is to maintain good health of our cars and ensure customer safety. Given the current wave of digitalisation sweeping across the nation, a lot of data regarding the working of vehicles is easily available. However, to process this data, we require an advanced AI-powered engine. This is also called upon because a majority of driver evaluation systems try to ape the West. However, given such a clear distinction in the driving styles and conditions, systems of the West is not applicable in India. Thus, riding on the back of superior technology made available in India, we see some futuristic players in India venture in the field of creating AI-based solutions.
“Riding on the back of superior technology made available in India, we see some futuristic players in India venture in the field of creating AI-based solutions”
Tell us about the technology behind the approach.
Fundamentally, two technologies have laid the groundwork for this approach. First is the Internet of Moving Things (IoMT), which makes it easier for companies like us to collect data about the vehicles movements and whereabouts in real-time. Second technology, of course, is AI powered dashboard cameras, which makes it easier for us to process real time driving behaviour and reveal actionable insights. When this advancement is coupled with OTA updates and notifications, we are looking at an integrated and seamless infrastructure.
Computer vision plays a major role in this automobile revolution; for instance, using computer vision to give real time collision warnings and other alerts. Please elaborate on this.
Certainly! We have come to live really connected lives these days. Our phones are connected with the self-drive cars or cabs that we hail. An engine that is able to process a wealth of information, made available through IoMT & AI, in real-time, is also equipped to send real-time notifications to drivers. Thus, if someone is driving at high speed and a pedestrian tries to cross the lane, our system alerts the driver in real time to mitigate mishap. This technique continuously analyses the live video stream from dash cam using a NVIDIA processor, detect objects, calculates probability of collision basis proximity and speed, thus, generating alerts and possibly saving lives. Using IOT data relayed from OBD device, we also detect rash driving events and calculate a driver score to give real time behavioural feedbacks. Our algorithm can also discover that the car may soon run out of fuel and accordingly alert the driver along with a list of nearby petrol pumps. We are sitting amidst the most exciting times when the industry is moving towards a revolution 4.0.
Zoomcar is creating India's first indigenous Driver score to encourage better driving. Tell us about this development.
We have launched India’s first vehicle model agnostic Driver Score Tech Stack for the passenger car segment. The AI Powered algorithm with machine learning capabilities tracks the mechanical specs of the car being driven, driving style of the customer and identifies critical events of driving and rates it on a scale of 0-100. The scoring system has capabilities to give real time feedback to drivers in the advent of rash driving to help them adjust their behaviour accordingly. Within the first month of its launch, the driver score has successfully reduced accident rate by 20 percent and maintenance and servicing cost by 25 percent for Zoomcar.