By Ariel Grossman, NoCamels -
AI cuts down process that can take days or weeks to a matter of minutes.
Taking your car to the garage for repair can be a frustrating process, but an Israeli company says it has made it possible for drivers to analyze their vehicle’s condition themselves – using just their smartphone.
All they need to do is open their camera, and start scanning their car. The AI, which company Ravin AI claims was trained with hundreds of millions of samples of car damage, can identify if a car has a dent or a scratch, and the extent of the damage of the exterior and interior. It will then generate an inspection and condition report of the vehicle.
“Every time you go and repair your car, or you claim against your insurance policy, or you’re trying to sell a used car, or you’re trying to buy a used car, there is a lot of ambiguity,” says Eliron Ekstein, CEO and Co-founder.
“There isn’t great clarity about the actual condition of the vehicle, which has been a consistent problem of the industry for years. And it comes from the fact that most people don’t really understand how cars work, and what’s the value of damages that they incur.”
Before founding the company in 2018, Ekstein was an entrepreneur in the auto and energy industry for a decade, and was the CEO of a startup that used machine learning to help taxi and Uber drivers find their next passenger by predicting demand patterns in local areas.
He got the idea for Ravin when he joined Shell as Head of New Business Development, to help the oil giant diversify from petroleum into services that could benefit fleets and drivers.
“We analyze the vehicle’s condition, such that both parties understand what’s at stake, and what kind of damages and issues the vehicle has. And that contributes to a much more clear understanding of the transaction and the right price to be paid.”
Ravin’s AI is a software as a service (SaaS) platform, in the form of a mobile-web app. Rental and used car fleets, car dealers, and insurance companies in the US and Europe sign up to its service and pay based on their volume of usage. The service is used by workers in these industries thousands of times every month – and their customers are able to access it through them.
It uses a number of patented technologies to create a model of the vehicle it is scanning on the spot. It understands what part the driver is looking at as they move around the car, and maps it on the generated model.
It tells users in real-time if they failed to scan a part of the vehicle, or if they are moving too quickly. The AI also gives the driver suggestions as to whether to file an insurance claim, or repair the car themselves.
The AI was trained with industry knowledge – specifically with the help of Ravin’s lead automotive, who was the former chief vehicle inspector at Mercedes Israel. That’s how the company says it can provide recommendations that are akin to a professional car inspector.
However, it does have its limitations. It struggles with assessing reflective parts, says Ekstein. And if the lighting conditions are poor, or there are harsh weather conditions such as snow or heavy rain, it will be unable to assess the damage to the car.
The company’s partners include Toyota North America, and American car rental company Hertz.
Ekstein says that over a billion vehicular inspections take place every year in the rental, used car, and insurance industries alone – and that for the most part, they’re a waste.
“Think about a rental car company that can’t operate a car because someone returned it with a ding on the door,” he says. “They can’t rent it out because it’s waiting to be assessed, and it’s really just lying there.
“We were working with rental car companies like Avis at the Newark Airport (in New Jersey, USA). And we saw dozens of cars just parked on the side – some of them were perfectly fine, but they just had some sort of markings and were waiting for an inspector.
“These are cars that are sitting idle, which means that the industry is producing more cars than it needs, which means more resources being wasted. And then the cost to the fleets also means that you’re probably overpaying for your rental car.”
Beyond that, he says that many of these inspections are subjective. His company claims to have tested this with a damaged vehicle that was assessed by three different professionals, and each one gave a different result – for the same exact damage.
“There is a lot of subjectivity in the assessments that we think can be eliminated.”
However, this won’t completely take out the need for car inspectors, Ekstein explains. “The AI in our solution takes care of that long tail of simple cases. But when there’s a really complex problem, for example a hit on the bumper and sensors are being affected, and someone really needs to look at the car and assess the impact to the value, then it goes to that professional who has years of experience.”
As for its competition, Ekstein says that there are several players, but that they have created an app that allows drivers to take pictures of the damaged car. Carscan is an example of such an app. And Deloitte, provider of consulting, tax, and advisory services to select clients, has launched its own AI-based app as well.
He says that compared to its competition, Ravin’s equipment is 10 times cheaper and more scalable.
“In terms of the technology itself, we don’t see anybody that’s using the same technology, in the sense that you can do a completely free flow scan, and the intelligence would pick up the damage itself and analyze it itself.”
The company is currently improving its AI so it can adapt and provide a condition report on the batteries and external condition of electric vehicles, a market it has not yet tapped into.
Ravin AI is based in Tel Aviv, and has approximately 70 employees. It has raised $30 million to date from investors including Shell Ventures and FM Capital, as well as individuals like Rick Wagoner, the former CEO of General Motors.
Caption: Ravin AI offers an AI-enabled vehicle inspection platform that detects, analyzes, and records damage in cars - using a smartphone.
Courtesy PublicDomainPictures / Pixabay
