An intelligent virtual assistant (IVA) in a carmaker's app, with access to the car's own data, can turn a warning light into self-service: it explains the problem, guides the driver through simple checks and fixes, and books the car into a garage. Garages get fewer calls and demand they can plan for. But it is not plug-and-play: it needs legacy systems connected and a well-structured knowledge base.
Why does car customer support need to change?
The automotive industry is full of marvels of engineering, with a history of innovation from Henry Ford's assembly line to cars that will soon drive themselves. Yet for such a carefully designed product, its customer support far too often lags behind.
What does a car know about itself?
A great deal. Cars have passed data to their drivers since the 1930s, when the Hudson Motor Car Company installed the first warning lights, for oil pressure and the voltmeter. Today a vehicle can have as many as 200 sensors: rain and light sensors that control the wipers and headlights, mass airflow sensors that work out how much fuel to inject, and sensors in the engine that report faults.
That data is processed in real time, to keep the car running and to warn of problems before they become serious or dangerous. In a predictive maintenance system, machine learning uses past data to predict when a part is likely to fail, so it can be fixed before there is any risk of a breakdown. Some problems the driver may never need to know about: Tesla's AI-powered diagnostic system can find and fix issues remotely. The rest are where technology for the customer can make every driver safer.
What happens when a warning light comes on?
Nothing very helpful. Most modern cars show an error code on the screen and tell you to call your garage. So you call, explain, and book a time, on the phone you probably use as your calendar, for a mechanic to diagnose the car and then fix it. You still don't know what is wrong, or whether the car will make it to the garage.
An IVA could change the whole of after-sales support. Some people still call them chatbots, but that doesn't do them justice.
What can an IVA do for drivers?
An IVA with natural language processing can live in the carmaker's phone app. With access to the car's data, it takes a customer from feeling helpless about their car to what customers in every industry want: self-service.
The IVA helps with troubleshooting and basic diagnostics, guiding drivers through simple checks, explaining what it finds and pointing them to possible solutions. Minor issues can be fixed at home, with no visit to a garage. A car is complicated, but a driver guided through simple maintenance and fixes is safer and better served, and more loyal to the brand.
What does it do for garages?
Reminders, prompts and warnings can be linked to local garages, so a garage knows in advance how many customers are due an oil change, a battery test or a service, and can plan for demand through the year.
Housed in your brand's app, the IVA can also guide customers to your own services and book them in: a few clicks, rather than a phone call or an outdated website. The garage loses the admin, and its staff can look after the customers in front of them instead of answering calls and emails.
Why isn't it plug-and-play?
Because the industry's customer-facing and garage systems are legacy systems, and many need better architecture. Before an IVA goes in, carmakers need to be sure it can work with the CRM, inventory management and other systems, so it gives customers accurate, up-to-date information.
It also needs a well-structured knowledge base, which is harder here than in other industries: a car typically has over 30,000 parts, and is a highly intricate piece of engineering. Distilling that into a knowledge base will take a lot of work, but it will pay off for the business and the customer.
After-sales support is only one use: IVAs can help across the whole process, including marketing and sales. With the right approach, a carmaker can change our relationship with our cars and its brand. How 4 Roads builds intelligent self-service.
Updated on 19 September 2026.
