Autonomous vehicles: are our cars driving us to the supermarket anytime soon?
#automotiveAutonomous vehicles: are our cars driving us to the supermarket anytime soon?

Stating that the automotive industry has changed throughout its more than 100-year history would be quite an understatement, as manufacturers seemingly can produce technological marvels every other day. After all, we went from the first car barely having a single horsepower to cars pushing that figure to over 1,000. The aerodynamics, as well as the technology found inside of the cars, jumped leaps and bounds and now the industry is discussing how can it integrate and massively produce vehicles that would be able to drive themselves. The case is not clear cut, however, as there are regulatory, security, and consumer-related concerns that automakers have to figure out before our cars drive us to the supermarket rather than the opposite variant of today.

At the basis of it, according to the Society of Automotive Engineers (SAE), there are six levels of automation, ranging from no automation at all (Level 0) to a driver being able to not pay attention to the road (Level 4) to have no steering wheel at all present on the car (Level 5). The European Automobile Manufacturers Association (ACEA) defined three levels of automated driving: assisted, automated, and autonomous driving. Per the three definitions, the drivers must consistently monitor the road with the car providing steering, acceleration, and braking support, transfer the tasks to the driver with sufficient lead time if road conditions do not meet its operational design domain (ODD), or eliminates any need for driver input, respectively.

Perhaps the most famous brand within the sphere of autonomous driving is Tesla’s Full Self-Driving system, which has had the ambition to become a Level 5, per the SAE classification, and is currently still in its early development processes. Thus, the current Beta version of the software is only at Level 2, which still will rely on the driver to take over in case things go south immediately.

But what has prevented the United States (US)-based Electric Vehicle (EV) maker, or any other automakers, to enable software to fully take over cars on the road and allow us to read our favorite books or scroll our beloved social media apps?

Early steps of autonomy

Modern cars are equipped with a wide array of radars, including such basics as the auto rain sensors, those that power the blind spot monitoring system, to more advanced radars that have allowed vehicles to measure the distance, direction, and relative speed to enable the usage of adaptive cruise control. Manufacturers have even introduced automated parking systems as early as 2003, once Toyota offered the system on the Prius.

The feature, though, was very primitive at the time. It had no radar and no collision warning system, meaning that the driver was responsible for stopping the car in case there was a pedestrian or a vehicle in the way. With the feature exclusively available in Japan, it was excluded from crossing the Pacific Ocean, and at the time, Judith Lee Stone, the President of Advocates For Highway and Auto Safety, “at first blush, it sounds like we’re not ready for this,” she was quoted as saying by The Chicago Tribune. Still, it was the first stepping stone for some form of autonomous driving to have reached the public, excluding the numerous tests and projects conducted throughout the 20th and 21st centuries.

As carmakers continued to run their sprint of technological development, features such as the already-mentioned adaptive cruise control, lane-keeping assistance, and automatic emergency braking are almost – and the term almost is used loosely – standard on today’s cars, resulting in many vehicles being equipped with Level 2 autonomy. A step above would be systems that combine adaptive cruise control and lane keep assist, and following that is the automated lane-keeping system (ALKS), which has been approved in the European Union (EU) from January 2022. The United Nations (UN), via the United Nations Economic Commission for Europe (UNECE), laid out the regulations for ALKS in June 2021, enforcing them a year prior to the EU.

“In its current form, the regulation allows the use of ALKS for speeds below 60 km/h on motorways – a use case initially applicable to assist in traffic jam or other slow-moving traffic situations,” noted the press release by UNECE, announcing that the regulation has been extended to trucks and other commercial heavy vehicles, including buses. “Once activated, ALKS are in primary control of the vehicle. However, the driver must be in a position to respond to a takeover request from the system.” The US is also part of the commission.

Per the SAE classification, ALKS is a Level 3 system. The classification designated it as “Conditional Automation”, meaning that while the car is technically capable of driving you to the supermarket, you would have to be monitoring the road continuously and intervene if needs be. While according to TÜV SÜD, Germany and Austria-based independent technical system certification provider, the UNECE regulations mean that “OEMs and vehicle manufacturers can implement an automated driving system in series-production vehicles,” the question is what limits carmakers from going all-in with Level 4 or even Level 5 systems?

Software is the burden?

According to Elon Musk, the sometimes head-turning and prominent chief executive of Tesla, the software is a big limiter. At least that was his opinion in 2016 when he told Forbes that “full autonomy is really a software limitation,” adding that the “hardware exists to create full autonomy.” For Musk, the focus has needed to go towards narrow Artificial Intelligence (AI). Instead of being a jack of all trades, companies have to develop a master of driving a car, rather than an AI that is capable of doing everything and anything. Volkswagen, the world’s second-largest automaker, promised to bring Level 4 autonomy to the market in 2019.

In the press release from the year, the Wolfsburg, Germany-based company announced that it would establish Volkswagen Autonomy GmbH (VWAT GmbH) with headquarters in Munich and Wolfsburg, and two subsequent subsidiaries in Silicon Valley and China in 2020 and 2021, respectively. Furthermore, all of the Group’s companies will unite under the VWAT banner to pool their resources and make strides in the development of autonomous driving, and “continue to use synergies across all Group brands to reduce the cost of self-driving vehicles, high-performance computers, and sensors,” at the time commented Alexander Hitzinger, Senior Vice President for autonomous driving in the Volkswagen Group and Member of the Volkswagen Brand Board of Management responsible for Technical Development at Volkswagen Commercial Vehicles (VWCV). “We plan to start commercializing autonomous driving at a large scale around the middle of the next decade.”

However, according to Hitzinger, there are challenges across the whole auto industry that are associated with introducing autonomous driving to consumers. “High development costs, extremely high demands on sensor technology plus a lack of regulatory systems and heterogeneous regional standards," are some of the biggest challenges, as pointed out by the company’s executive.

Thus, concluding that software is not the only roadblock for autonomous vehicles would not raise too many eyebrows. But can the automotive industry overcome these challenges in order to allow consumers to sit in a vehicle that has no steering vehicle whatsoever?