AI: Do more than keep the vehicle in its lane

Everyone has heard the creaking or unusual noise from under the hood, and it’s too late, we can only watch the car steaming and parked on the side of the highway because of a breakdown. A survey by UK-based car repair company Kwik-fit shows that the average cost of a breakdown repair is around £350, and in some cases it can cost upwards of £1,500. A lot of faults that need repair could have been caught before they got to the point of failure, but there’s no way a mechanic is sitting in the passenger seat all the time, so how do we anticipate these problems and deal with them earlier? The answer is artificial intelligence (AI).

AI is not limited to driver assistance functions

From super cruise systems to autonomous driving systems and more, the fields of advanced driver assistance systems (ADAS) and fully autonomous driving have become the backbone of the automotive industry. Many of the latest and greatest AI systems are ready to be used in automakers’ highest-quality automotive products. AI can do more than help keep vehicles in their lanes, back up seamlessly, and direct you to your favorite coffee shop; AI is very good at recognizing patterns and predicting outcomes based on simple sensor data that might Too subtle to be perceived by human senses, making AI the perfect tool for capturing unusual sounds before they turn into “expensive” loud bangs.

AI needs to be trained on high-quality data in order to work efficiently. Fortunately, vehicle breakdowns, repairs and maintenance are well documented, while breakdowns recorded by mechanics and annual tests performed by departments of transportation alike can provide a wealth of empirical evidence of vehicle operation. Taking this data and feeding it into a machine learning algorithm can provide an efficient model that can predict possible failures based on data points ranging from exhaust emissions to slightly uneven brake decay.

Previously, we already had an omniscient and omniscient “mechanic” in the cloud, and now we need to deploy this intelligent function to the vehicle at the edge. The next step is to connect a very basic neural network accelerator (such as Imagination’s NNA) to the vehicle’s engine management computer and other existing sensors in the vehicle, such as tire pressure and temperature sensors. With a couple of additional sensors that measure things like abnormal vibrations, AI can spot potential failures before they happen.

That way, when the AI ​​detects a situation it thinks could soon turn into a malfunction, it alerts the driver of what it has detected and advises them to go to a mechanic for a quick, inexpensive repair, rather than sit back and watch It ended up being parked on the side of the road due to a malfunction. The system can also provide mechanics with more detailed information about perceived faults, freeing them up to provide high-quality repairs with less time and effort when troubleshooting.

Deploying this “AI mechanic” into every car, from the most luxurious car to the most basic short-haul city car, may sound like a lot of money, but know that when a breakdown occurs, it’s not just The client’s pocket.

Minimize the cost of failure

Frustration and anger are two of the most common emotions when a breakdown occurs, and some of these emotions are specific to the breakdown situation, but usually these emotions are specific to the vehicle and its manufacturer. On top of the hundreds of pounds it costs to repair the car, if the driver goes looking to buy a new car, it can be a huge loss to the original car brand because he has vowed never to buy your brand again, which means Your sales will do nothing to it.

There’s also the question of who exactly is paying for breakdowns. Vehicle ownership has changed dramatically, with more people adopting leasing, financing or loan-based ownership models than ever before. These ownership models often include extensive warranty and repair programs that, in some cases, span as long as 100,000 miles and 10 years.

In the eyes of consumers, a long-term warranty means that everything can be taken care of, and if something goes wrong, they automatically assume that the dealer or the automaker should pay for the repairs.

If AI systems are used to monitor the condition and likelihood of failure of millions of vehicles, manufacturers can spot potential system problems before they escalate into larger problems and proactively address them.

Fast-paced life

No one likes problems, and car breakdowns are no exception, whether for individual car owners or fleet managers worried about the impact of repairs and fleet downtime. By deploying simple, low-cost neural network accelerators and other computing devices, many headaches can be solved before they happen, turning the smoke from a car breakdown and a driver’s rage into a drive to the garage. A hassle-free journey, all it takes is a simple repair within the warranty period to keep the customer running.

Imagination has an industry-leading portfolio of Neural Network Accelerator (NNA) IP, which are purpose-built AI accelerators that deliver leading edge AI inference performance. Starting with the release of the 2NX NNA in 2017, Imagination has released three generations of NNA products. Among them, 3NX can help automotive applications to optimize computing power and performance; and the latest generation of 4NX is the ultimate AI acceleration solution specially designed for ADAS and autonomous driving applications. In terms of safety, it can also help customers obtain ISO 26262 certification faster with the design process in compliance with the ISO 26262 standard.

The Links:   EL480240-PR1 7MBR50UA120-50 IGBTMODULE