The ongoing debate surrounding Tesla’s Full Self-Driving (FSD) capabilities has taken an interesting turn with the introduction of a new test involving a fake wall. This test follows a controversial video by former NASA engineer Mark Rober, where a Tesla Model Y drove through a painted wall, purportedly while using FSD. The purpose of this experiment was to challenge Tesla’s vision-only driver assistance system, which relies solely on cameras for obstacle detection, unlike many competitors that incorporate radar and LiDAR technology.
Rober’s original test drew scrutiny for using basic Autopilot rather than the full FSD and for raising questions about the authenticity of his methods, especially with the presence of a LiDAR-equipped vehicle from Luminar. This led to skepticism regarding the objectivity of his findings.
In a bid to obtain clearer results, content creator Kyle Paul conducted a straightforward experiment using two Tesla electric vehicles (EVs) to replicate Rober’s test. This new approach is significant, particularly for owners of older Tesla models, as the performance of Autopilot and FSD varies significantly across different generations of vehicles.
Filmed on a closed course, Kyle Paul enabled FSD for all attempts. The first vehicle tested was a Tesla Model Y equipped with the previous-generation Hardware 3 computer running FSD version 12.5.4.2. While this version is not the latest—12.6 being the current iteration—it was deemed sufficient for this particular test. Unfortunately, the Model Y failed to detect the fake wall, requiring the driver to intervene and brake manually on multiple occasions. This was particularly concerning given that the fake wall was reinforced by a truck, which could have resulted in severe damage had the vehicle collided with it.
After the disappointing results with the Model Y, Kyle Paul introduced a Tesla Cybertruck into the test, which boasts the latest Hardware 4 and is running the newer FSD V13 software. The difference in performance between the older Model Y and the cutting-edge Cybertruck was immediately apparent. As the Cybertruck approached the fake wall with FSD enabled, the obstacle was detected early on the vehicle’s touchscreen, and it came to a complete stop autonomously—demonstrating a significant advancement in Tesla’s technology.
This comparison highlights the ongoing evolution of Tesla’s autonomous driving capabilities, particularly with the integration of advanced hardware and software in newer models. While the likelihood of such a scenario occurring in real-world conditions is slim, it raises important questions about Tesla’s commitment to using only camera-based systems for its driver assistance features. The improvements seen in the Cybertruck suggest that Tesla is making strides in enhancing obstacle detection capabilities, although older vehicles with Hardware 3 may not see these benefits immediately.
In summary, the contrasting performances of the Tesla Model Y and Cybertruck in Kyle Paul’s fake wall test underscore the advancements made in Tesla’s autonomous driving technology. As the company continues to refine its FSD capabilities, it remains to be seen how older models will adapt to these innovations. While the debate over Tesla’s reliance on camera-based systems persists, the results from this experiment suggest that significant improvements are possible with newer hardware and software configurations.
