First Level 4 Autonomous Car in the EU Without AI: How Is That Possible?

What Happened?

Aidoptation, a Belgian tech company founded in 2020, has obtained the first Level 4 permit in the European Union to test an autonomous car on public highways. The authorization, granted by the Belgian transport agency FPS Mobility, covers 100 km of the E313 and E314 highways in Limburg province, Belgium. This allows the vehicle to operate without human intervention at speeds up to 120 km/h, even without a driver on board. Most notably, the driving system does not use artificial intelligence, but rather an approach based on sensors and deterministic logic. According to The Next Web, the company announced the approval on July 15, 2026, highlighting that it is the first Level 4 permit in the EU for public roads open to general traffic.

Why Is This Important?

This milestone challenges the dominant narrative linking autonomous driving to machine learning. By forgoing AI, Aidoptation avoids the black box, bias, and lack of explainability issues that plague other systems. This could accelerate regulatory approval, as deterministic algorithms are easier to audit and certify. Additionally, it demonstrates that Level 4 is achievable with simpler and potentially safer technologies. In a context where accidents involving AI-driven autonomous vehicles have generated distrust—such as the case of an Uber that struck a pedestrian in 2018 or investigations into Tesla's Autopilot system—Aidoptation's approach offers a transparent alternative. Explainability is key to public and regulatory acceptance, and this system could set a new standard.

Consequences for the Market and Regulation

Aidoptation's approach could pressure other manufacturers to reconsider the exclusive use of AI. Companies like Waymo, which has accumulated millions of test kilometers with deep learning, or Tesla, whose Full Self-Driving system relies on neural networks, could face additional scrutiny if their systems cannot explain their decisions. In the EU, the General Safety Regulation (GSR) requires autonomous vehicles to be safe and traceable; a deterministic system easily meets these requirements. Moreover, the Belgian decision sets a precedent for other member states to harmonize their regulations. According to an Aidoptation spokesperson cited by The Next Web, "This approach demonstrates that AI is not the only path to autonomy. Safety and transparency can be equally important." In the long term, this could foster more diverse competition in the autonomous driving market, where startups with alternative technologies have a place.

What Readers Should Know

Aidoptation's car uses a combination of lidar, radar, cameras, and high-definition maps, processed through predefined logical rules. It does not learn from experience but follows programmed instructions. This limits its adaptability to unforeseen situations, such as roadworks or extreme weather, but guarantees predictable behavior. The test in Belgium is a first step; commercialization is still far off. The company has not disclosed the vehicle manufacturer or technology partners, though speculation suggests it could be a modified production model, such as a BMW or Volkswagen. The business plan has also not been detailed: will they sell the system to manufacturers or operate it as a mobility service? What is clear is that the permit is valid for one year and requires remote supervision in case of emergency, though without the need for a driver on board.

"This approach demonstrates that AI is not the only path to autonomy. Safety and transparency can be equally important," says an Aidoptation spokesperson.

Historical Context

Until now, Level 4 autonomous vehicles in Europe had only been tested in controlled environments or with remote supervision. For example, in 2023, the German company Vay tested a remote-controlled car in Hamburg, but it was not fully autonomous. In 2024, Waymo conducted tests on California highways with safety drivers. This permit is the first to allow operation without a driver on board on public roads open to general traffic in the EU. FPS Mobility's decision was based on technical audits that verified the safety of the deterministic system. Belgium thus positions itself as a mobility innovation hub, similar to what Germany did with its 2021 autonomous vehicle law, which allowed Level 4 tests in designated areas. However, Aidoptation's authorization goes a step further by not requiring human supervision on board.

Comparison with Other Players

While Waymo and Cruise have conducted tests in the US with AI, Aidoptation takes a different path. Its system is more similar to Level 2+ autonomous vehicles that use adaptive cruise control, but extended to full highway driving. The absence of AI reduces computational and energy costs: a deterministic system can run on more modest hardware, making the vehicle cheaper. In contrast, Waymo's systems require powerful computers and consume a lot of energy. Additionally, the lack of AI avoids bias issues in decision-making, such as those affecting pedestrian recognition systems. However, limited adaptability could be a disadvantage in complex urban environments. Aidoptation initially focuses on highways, where conditions are more predictable, which is a smart strategy to demonstrate viability.

Speculation and Unconfirmed Aspects

It is unclear whether Aidoptation plans to expand tests to other roads or countries. The vehicle manufacturer and technology partners have also not been revealed. Speculation suggests it could be a modified production vehicle, but there is no official confirmation. Additionally, the cost of the system and whether the company has agreements with insurers or manufacturers are unknown. Another unconfirmed aspect is the system's ability to handle emergencies, such as an animal crossing the road. Aidoptation claims the system can brake or avoid obstacles, but there is no public data from exhaustive tests. Finally, the reaction of European regulators may vary: while Belgium has been a pioneer, other countries like France or Germany might require more tests before authorizing similar permits.

Conclusion

Aidoptation's permit marks a before and after in European autonomous mobility. By demonstrating that a system without AI can achieve Level 4, it opens the door to alternative approaches that prioritize explainability and safety. The coming months will be key to seeing if this technology can scale and compete with AI giants. If successful, it could accelerate the adoption of autonomous vehicles in Europe, reducing costs and increasing public trust. However, questions remain about its adaptability and commercial viability. What is certain is that the race to autonomy is no longer a monopoly of artificial intelligence.