Email
Password
Remember meForgot password?
    Log in with Twitter

article imageQ&A: Using AI to detect vehicle threats and wear Special

By Tim Sandle     Dec 25, 2018 in Technology
The automotive industry is undergoing major growth and transformation. Among the innovations is the use of AI to assess vehicles for both threats and performance. Amir Hever of UVeye explains how.
As global auto sales continue to rise worldwide (led by emerging markets), dozens of companies - from traditional car manufacturers such as Audi, Ford, Volvo, BMW, as well as tech brands such as Intel, Google and Microsoft, are investing in the autonomous trend.
According to Amir Hever, CEO and co-founder of UVeye, as more vehicles make it onto our roads, the more critical it is to identify anomalies (including wear and tear, and even security threats such as bombs) to ensure vehicle safety.
UVeye is applies computer vision and machine learning technology to conduct a 360-degree scan of a vehicle in high resolution in three seconds, even when the car is traveling at speeds of up to 28 mph (45 km/h). Digital Journal spoke with Amir Hever to discover more.
Digital Journal: What is the state of the global automotive market?
Amir Hever: Like many industries, the automotive market is undergoing a tremendous shift, driven by technology. We’re seeing a surge in demand for hybrid, electric and autonomous vehicles, which will fundamentally change the automotive landscape in the near future. As a wider variety of vehicles make it onto our roads, this will lead to more questions about how human-driven cars and autonomous vehicles can co-exist, what regulations need to be put in place, and what systems and processes need to be enacted to ensure all vehicles are in a safe and good condition at all times.
DJ: How big will autonomous vehicles become?
Hever: The global autonomous vehicles market is undoubtedly booming, with a recent study finding that driverless vehicles will be responsible for $7 trillion worth of economic activity and new efficiencies by 2050.
While self-driving cars are not yet ready for prime time - and consumers are still generally hesitant to ride or drive these cars - we’re seeing steady progress from a variety of companies to make autonomous vehicles a reality. This month, for example, Waymo (a Google spinoff) launched its first autonomous ride-hailing service in Phoenix, Arizona. Called “Waymo One”, the robot taxi service is exclusively available to a few hundred riders. Waymo is just one of the many companies expected to roll out autonomous car fleets in the near future.
DJ: Which types of companies are investing in autonomous vehicles?
Hever: Autonomous vehicles are taking the majority share in auto tech funding, with 70 percent of the total $5 billion invested this year. While a slew of venture capital funding is being injected into auto tech, virtually every traditional automotive manufacturer is also aggressively investing in autonomous innovations.
For example, BMW, Intel and Mobileye have forged an alliance with the aim of launching a self-driving car that will be ready to drive by 2021. Meanwhile, Toyota announced a $1 billion budget for autonomous driving research, working with professors and researchers from Stanford University, MIT and the University of Michigan. Tech titans such as Amazon and Apple also have teams dedicated to driverless vehicle technology.
DJ: How important are vehicle safety assessments as vehicle technology advances?
Hever: As more cars - and car types - make it onto our roads, the more critical it will be to have a solution that can immediately identify any threats or anomalies, as opposed to waiting for periodic checks.
The next wave of automation is taking place in the automotive and broader transportation industry, and vehicle inspections need to move away from traditional time-consuming, low resolution and error-prone processes. At UVeye, we remove the guesswork by conducting a holistic vehicle security and safety inspection throughout a car’s entire lifecycle, saving time and money for our customers.
DJ: Why do vehicles safety issues still occur?
Hever: Vehicle safety issues still occur because the process for conducting inspections remains a very manual and tedious process, prone to human error. There are two key moments when a vehicle needs to be assessed. The first is during the assembly line, checking for everything from faulty brake systems, stalled engines to other potentially dangerous issues. Unfortunately there have been cases where alleged improper testing of vehicles has led to recalls. At UVeye, we are able to mitigate such issues by not only assessing individual parts at the assembly line, but also when the automotive is assessed as a whole at the end of the manufacturing process.
The second instance when vehicle safety issues can occur is through general wear and tear. While oil leakages and visible damage such as dents and scratches can be easily located, inconspicuous wear and tear issues are much harder to detect. Rather than dropping off the car at a mechanic - often for days - to have it checked out, UVeye can scan the vehicle in seconds to determine any major problems. Since we can assess a car throughout its entire lifecycle - from production to when it’s taken off the road - we can proactively identify and report on any problems with specific models.
DJ: How can vehicle recalls be reduced?
Hever: Recalls can be reduced by conducting regular, instant and accurate inspections of the vehicle throughout its entire lifecycle. In the future, as demand for cleaner vehicles soars, we predict that conventional gas stations will move towards more maintenance oriented-services - including offering vehicle charging and smart inspections.
We’re already seeing this happen - for example, BP and Royal Dutch Shell have already invested in electric vehicle charging companies, and have rolled out units at retail sites. UVeye could be implemented in these smart maintenance stations, as well as other venues where vehicles can drive in to do inspections instantly.
DJ: What services does UVeye provide?
Hever: UVeye is an AI-powered vehicle inspection and security system that identifies threats or anomalies to the vehicle’s exterior. It detects these issues by conducting a 360-degree scan in seconds, which can be done anytime in a vehicle’s entire lifecycle. These potential issues can include oil leaks or rust, dents and scratches or more malicious threats such as hidden explosives or drugs.
UVeye’s solution can scan vehicles up to approximately 11.5 feet in width, with no limitations on its length. Using multiple strategically angled and synchronized cameras, we are able to generate a 3D, high resolution and full color image of the vehicle - even when the car is moving at speeds of up to 28 miles per hour. Our proprietary technology compares the image to vehicle manufacturer-supplied data, which can accurately identify any threats or anomalies throughout a vehicle’s entire lifecycle - starting from the assembly line to when it has been taken off the road.
DJ: How did you develop your technology?
Hever: The idea for UVeye came from my experience driving into a government facility one day. As I drove towards the entrance gate, a security guard knelt down to look underneath the vehicle to inspect for any threats. I recognized that this was a huge flaw in this security process and developed the technology to more effectively and quickly spot any security or safety threats or anomalies.
At UVeye, we’ve honed our technology over the last two years, bringing onboard seasoned experts in AI (deep learning and machine learning), computer vision, security and other disciplines.
DJ: Which companies do you work with?
Hever: UVeye’s customers span two key areas: security and the broader automotive/transportation industry. We initially started working with organizations focused on securing their offices and perimeters, such as government agencies, embassies, port facilities and other critical infrastructures such as power plants and hotels.
We soon realized that our technology could benefit the broader automotive/transportation market, which can include car rental companies, dealerships, vehicle repair shops and any company operating a large fleet of vehicles. Our customers include Toyota Tsusho and Czech auto manufacturer Škoda (part of Volkswagen Group), and other automotive manufacturers. Today, our systems have been installed worldwide.
More about Automotive, Cars, Artificial intelligence
More news from
Latest News
Top News