Why is Tesla developing an autonomous car

Autonomous driving: Tesla's accident and the industry's major hurdles

The autopilot was probably switched off. Still, after the death of two passengers in a driverless Teslahat, attention is once again drawn to a market notorious for over-promising and under-delivering. Five or six years ago, every company - from the then newcomer Uber to the vintage Ford - had promised to put self-driving vehicles on the road in the 2020s. The deadlines have already passed. This is probably one of the reasons why the companies have become somewhat more modest in terms of their ambitions.

When will the breakthrough come?

The market for this technology is as complex as the vehicles themselves. There are many interlinked parts, all of which are at different stages of development. That makes it difficult to say if and when self-driving vehicles will finally make the breakthrough. The technologies for long-haul trucks, light private vehicles, taxis and other forms of public transport require different approaches. And so far there are only a few self-driving cars on the road. This is why the companies still receive little data. That also slows the progress of the industry.

Autonomous driving - here a self-driving car (Volvo) from Uber in Washington (2020)

"The functionality, the software and the algorithms that are required are all based on data. And the database will grow over time," said Alexander Petrofski, chief strategy officer at Volvo Cars, told Bloomberg. "Unfortunately there is no switch for this, but we see a scenario in which we gradually get more information from more regions." However, the Swedish car manufacturer has postponed the debut of its autonomous vehicles to 2022.

The approval requirements are just as complex. The regulations governing the conditions under which an autonomous vehicle may drive on public roads vary not only by vehicle type, but also from country to country. And often also on a national and regional level.

Who is in the lead?

The industry has now grabbed everyone: from old car giants like Ford and Volkswagen to small robotics startups in Israel and the USA. In the US state of California alone, more than 50 companies currently have a license to test autonomous vehicles on the road with a safety driver. Experts assume that the market will increasingly consolidate as progress continues.

Waymo self-driving car

Waymo, owned by Google, is considered by many to be the most advanced when it comes to autonomous driving. It has been operating a fully driverless commercial taxi fleet in Phoenix, Arizona since October 2020. But the company is still years away from being profitable. In 2018 it announced plans to expand its taxi fleet to "up to" 82,000 vehicles. Today the network has a few hundred cars - probably a sign that something has not gone according to plan.

Other companies are also following the Waymo strategy and cooperating with local taxi services. The support of the big automakers could give them an edge when it comes to expanding their business model quickly. Cruise is supported by General Motors and Honda, Argo from Ford and Volkswagen and Motional from Hyundai - they all follow the model of a taxi fleet.

Is Tesla falling behind?

Elon Musk's Tesla sells its cars directly to consumers. According to the Tesla homepage, the built-in autopilot enables ″ autonomous driving today ″. As a result, the company already has a wealth of data from the real world. But the autopilot is best known for fatalities.

"Tesla is falling behind other automakers like GM and Ford, who are using technology in models with driver assistance systems that ensures that the driver is looking at the road," Jake Fisher, an expert on car testing for Consumer Reports, told Reuters after the recent accident .

In order to compete with Google, Amazon also entered the race for self-driving cars last year and acquired the startup Zoox. In December, tech startup Aurora bought Uber's self-driving division after the company sold off the unprofitable project. Former Uber CEO Travis Kalanick had earlier predicted that the company's human drivers would largely be replaced by automated vehicles by 2018.

"In the race to commercialize autonomous driving, many players will not make it across the finish line," Aurora CEO Chris Urmson told Forbes in December, saying the company has a chance of being one of the winners.

A number of companies are also taking part in the race in China: names such as AutoX, WeRide, Pony.ai, Baidu and Didi Chuxing are working on self-driving cars. In December some of them were allowed to take to the streets in China for the first time - albeit only under test conditions, which suggests that the market in China is not yet particularly advanced.

The problem with energy efficiency

The developments in the field of e-mobility are still challenging the developers of autonomous cars. In order to navigate independently, many self-driving vehicles rely on lidar sensors - short for light detection and ranging. This is a sensor on the roof of the vehicle that emits laser waves and thus determines the distance to surrounding objects. The position of the sensors increases the wind resistance of the vehicles and this influences the energy consumption.

A few companies, including Tesla, have replaced lidar sensors with cheaper camera technology. Here are the cameras behind the car windows. The energy problem can be avoided here - but they are considered less precise. However, many in the industry also assume that the issue of energy will soon be able to be dealt with. As a result, the vehicles could then also fully jump on the e-mobility train. However, one should still be careful when estimating when the autonomous cars will hit the streets in large numbers.

Adapted from English by Nicolas Martin.