Why data handling may put a bump on the road to autonomous driving

Autonomous driving is a technology of the future. If one day vehicles will be able to drive themselves, the concept of what a “car” is will need to be rethought.

But, there is still a long way to go. Handling the amount of data required by autonomous vehicles and the prerequisites that this data must meet are important, but so too are the vehicle’s autonomous safety measures.

If self-driving cars are on the roads one day, they will have to process very different data types and formats in real time.

Data comes from the vehicles themselves as they perceive their surroundings through cameras and various sensors, to decide whether there is an object on the street that they have to avoid.

It also comes from the vehicle’s own status, such as if the tank or battery is almost empty. Then, they must be able to make the decision to go to the nearest fuel or charging station.

Autonomous cars will also record and process information from their surroundings.

This can help to increase road safety by, for example, other vehicles sharing information about impending traffic jams or changing weather conditions.

In so-called “smart cities,” it may even be possible for vehicles to exchange information with smart roads or traffic lights.

Finally, the cars also collect data about their owners and use it to make decisions.

This may be the preference for road trips instead of motorways, a preferred route to work, which may not be the fastest, or the supermarket where shopping is always done.

One of the biggest advantages of self-driving vehicles is to increase road safety. While people are distracted or tired by external influences, this does not happen to machines.

Your decisions are based on learned knowledge and real-time data processing.

The data sets, which are used to train AI systems of vehicles and are intended to guide their decisions in traffic, must accordingly be of very high quality, otherwise accidents can also occur in self-driving cars.

If two passers-by and a parked car can be seen, this information must be correctly noted. If not and the AI learns there is only one pedestrian and a car, this can lead to misinterpretations later on the road which could have fatal consequences.

The same applies to data that vehicles receive and process in operation, so everyone from  automakers to network infrastructure operators must ensure that the data is as complete and trustworthy as possible.

To help build the confidence that consumers need to accept and use self-driving cars, the vehicles must also be protected from hackers.

Already, connected technology in cars offers criminals various entry gates, be it via Bluetooth modules or keyless systems.  

In the case of self-driving vehicles, the variety of functions will continue to grow, which will also increase vulnerabilities.

There is a danger not only from attackers who want to cause accidents, but also from those who want to steal data.

The development of autonomous vehicles is much more complicated than that of conventional cars. Automakers have to solve much more complex tasks in the software sector than ever before.

Undoubtedly, the automotive industry will undergo far-reaching changes in the coming years and decades, and only automakers that adapt to these changes will succeed in the long term.

So, they need to develop new business models to generate profits.

Today, there are the first examples of auto companies generating revenue by unlocking certain functions, such as high-beam assistants or adaptive chassis, for a fee.

By combining the status data of the vehicles, the owner’s behavioral data and the environmental data in a central platform, they can analyze all the data available to them, to identify other new products and services they can offer to their customers.

File source

Show More
Back to top button