Is Automotive Ready for AI?

Automotive manufacturers, vehicle insurers and maintenance-repair operations all seek to leverage real-time information. But first they must transcend technical limitations and data integration difficulties.

Companies that leverage artificial intelligence say AI and cloud technology can connect data from those sectors. Though it’s a widely used term now, they say its capabilities weren’t available in automotive retail until fairly recently, say Sumit Chauhan, co-founder and chief operating officer of CerebrumX, and Aharon Horwitz, CEO and co-founder of Fullpath, an AI-powered customer data and experience platform.

AI technology took a quantum leap recently, Horwitz says, and caught car dealers’ attention, “and now they are excited about it and its possibilities.”

Outside automotive retail, the AI road is already full. CerebrumX, for example, is working on an application that gathers information as people drive. Some vehicles come equipped with a configuration of radar, laser-emitting lidar, and cameras to paint real-time pictures of roadways, traffic and other conditions. In the application, AI can inform drivers about the things they see on either side of the road. For example, the building on the right is a historical building and there is a historical marker ahead. “All this information can be delivered conversationally,” Chauhan says.

Traditionally popular FM radio also provides interesting AI use cases. Most U.S. drivers, according to Edison Research, still listen to FM radio in their cars – 73%, in fact. Here, AI is poised to provide information on specials at local service shops and businesses. For example, Starbucks might place an ad that plays shortly before a vehicle passes a Starbucks.

Another potential application would give insurers and service shops usage data and demographic information to assess risk. “Today, insurers ask for a driver’s license, name, age, ZIP code, education level, marital status and number of drivers,” he says. “But what if insurers could base insurance policies on the last three to six months of driving behaviors through historical data? This would allow insurers to develop better risk profiles and price policies accordingly, and also pass the benefits back to the end users.”

Tire companies and repair shops could also access AI to determine a vehicle’s repair needs. For instance, newer vehicles’ tires are equipped with sensors that gather and relay data on friction and road conditions and hence wear and tear.

“Some of these use cases were not available earlier because we lacked access to that data,” Chauhan says, explaining that before the applications can become commonplace, companies must learn to power the process and integrate the information available from data-collection sources on the vehicle.

Insights for Insurance

If insurers could base insurance costs and policies on more detailed driver profiles, it would benefit all involved, Chauhan says.

Accessing information collected by AI telematics vehicle systems and other autonomous technologies would give insurers data on acceleration, braking, cornering, collision detection, lane-change direction and more. With this information, they could price policies based on individual risk profiles, giving back 10% to 15% from premiums to consumers with safe-driving habits.

“AI could enable more viable unit economics for the insurance industry because as costs keep rising and vehicle prices soar, insurance premiums are going up, as well,” Chauhan says.

Chauhan cautions that the capability doesn’t mean insurers have unfettered access to the personalized data but can access it only if a consumer consents in order to get a usage-based policy.

Online Assistance

AI can also address online communications, taking chatbots to the next level, according to Horwitz.

Dealers can load their vehicle inventories, special discounts and other promotions into a chatbot so that when the consumer asks a question, the system responds with data about current inventory, specials and offers, service hours, and more.

For instance, if a customer tells a chatbot he’s interested in a blue Camry, it can respond with information about that particular vehicle in the dealer’s inventory, plus details on any specials running on that model.

“In every single conversation, the system will give the customer the exact information they are looking for,” Horwitz says, explaining that targeted conversations can lead to higher conversion levels. “They receive quality information in a conversation that answers all their questions. It’s a good experience that encourages them to come to the dealership.”

With AI, a dealership also will never call again, adds Devin Daly, CEO and founder of Intel, an AI provider that works with about 5,000 dealerships in 53 countries. Leads typically come in after-hours. Here, an AI-powered solution could lead the conversation, continue the engagement, and answer specific questions. When the sales team arrives the next day, a customer is prepared for them to approach.

Veer into Vehicle Service Warranties

Administrators of vehicle service warranties also benefit from AI technology, according to Chauhan.

When car owners take out extended warranty policies as their vehicles fall out of the original warranty, they can consent to share information about the car and its overall vehicle health via AI. In the past, that information came from service provider maintenance records.

 “AI can give more color and three-dimensional insight into actual vehicle health,” Chauhan says. “Now we not only have access to maintenance records but to actual vehicle health information, such as battery health, fuel pressure levels, tire pressure levels and how many times the check engine light came on in the last six months.”

F&I can use the information to make better decisions on extended warranties, he says, including whether to offer them and at what price. “Once the extended warranty is in place, they can monitor the car’s health in real time.”

Again, that application can be employed only if the consumer consents to sharing their data. Their consent, however, benefits them as well, Chauhan says. If problems with the vehicle arise, dealerships can use that information to communicate with end users about vehicle repair needs. “This way they can take the vehicle to the service station to avoid nullifying their warranty.”

Service stations and dealership service operations also benefit because they can learn when tires or batteries need replacing, then dispatch targeted marketing offers to customers for tire rotations, replacements, vehicle alignments and more.

“Access to this data allows for a very pointed marketing strategy that can provide instant and additive revenue to these companies,” Chauhan says.

Push for Predictive Maintenance

Predictive maintenance techniques help determine vehicle condition to estimate when maintenance should be performed. This approach promises cost savings over routine or time-based preventive maintenance because tasks are performed only when warranted.

Predictive maintenance can benefit fleets by helping them optimize and intelligently plan how to allocate resources, which can minimize maintenance expenses. It allows fleets to plan downtime when it’s convenient and ensures vehicles are kept in tip-top condition.

“It is extremely important for fleets to know the exact health of their vehicles and when it is time to make repairs or even sell them,” Chauhan says.

Consumers can also benefit from the data. “There are a lot of grease monkeys who want to service their own vehicles,” he explains. “Predictive maintenance data has immense value here. It gives these owners insight into what is wrong with the vehicle, so they can decide whether to fix it themselves.”

Focus on Electric Vehicles

Electric-vehicle batteries have multiple cells generating information every second. A Tesla, for example, produces about 100 values per cell, per second, making for much complexity in determining whether something’s amiss with the battery, Chauhan says.

But the data packs much opportunity for service operations, he says, “if someone can provide a platform to do that in real time and across OEMs and battery manufacturers. That’s one focus area for us. How do we monitor the actual battery health inside the auto?”

The second focus is monitoring battery strength as the vehicle moves down the road. Chauhan says AI can ease drivers’ anxiety about whether they have enough range to reach the next charger.

“Drivers must turn off excessive lighting, stop charging their phones or turn down air conditioning to extend battery power ... We have algorithms that can help the end user make decisions in real time to reduce their anxiety.”

The third EV area AI can help with is in the marketing of dealership charging stations.

“Businesses can send out targeted offers to get drivers to come to your charging station, where they can charge their vehicles at attractive rates or with a membership,” Chauhan says. “This could lead to a subscription model. We are working with infrastructure providers to deliver insights on how they can optimize their businesses and increase revenue by attracting customers to their charging stations.”

Risky Business?

With every innovation, there’s also risk, including with AI, Chauhan acknowledges. “Automotive AI systems are more open to the world, making these concerns greater.”

The systems must therefore produce solutions that protect data privacy and security breaches, he says. “I am confident the industry will do this, but the level of investment and commitment toward connective vehicles is just reaching an inflection point. The issue with data privacy is just now coming into focus.”

The other risk is regulatory, he says. “There are governments who want to democratize the data,” he says. “They want to collect the data and give it back to the people. That’s a very humanitarian idea but not how business works. The day you tell someone to do something for free is the day they will stop doing it properly. Pushing OEMs to democratize their data will stop innovation in this space.” On the other hand, he says governments should play a role in getting AI technology rolled out.

“A level of acceleration is also needed in the chipset industry. The chipsets are very expensive, and there are supply chain issues. There is a lot of innovation right now from existing and new players. Those innovations should address these concerns.”

Daly identifies another concern—this time with employees who fear AI will take over their jobs. It’s a legitimate fear if workers do not understand the technology is meant to help not take over, he says.

“AI is your superpower,” he adds. “It’s an assistant that never takes breaks, works 24 hours a day, every day,” he says. “It can handle tedious mundane tasks to keep employees building relationships and closing deals.”

Integration of various data sets is the key to making it work and reaping its true benefits in automotive.

“AI is going to unlock a new level of productivity for auto retail,” Daly says. “And it will get rid of the mundane tasks no one likes to do. But to be successful, work with a vendor who knows what they are doing, trained to provide a specific experience designed for the automotive industry. That will help you achieve the results you’re looking for.