The Paris Motor Show test drives offered a striking demonstration of how far supervised driving assistance can go in busy city traffic. But Momenta’s intervention-free ride, Tesla’s brief test and Xpeng’s cautious speed choices were observations from different drives—not a controlled contest or proof that any system is ready to drive without supervision.
5
7
43
What the three Paris drives showed
Momenta R7 in Chery’s Jaecoo 7: In Chunying Zhang’s reported drive, the Jaecoo 7 spent about an hour in Paris traffic without a driver takeover, including the Arc de Triomphe roundabout. The junction is especially challenging: 12 avenues converge there, lane markings are unclear, and entering vehicles have priority. The car’s ability to navigate the roundabout and choose an exit was an impressive demonstration. But abrupt braking and a last-second maneuver around a blocked lane also showed that completing a route without intervention is not the same as consistently smooth, anticipatory driving.
5
7
43
Tesla Assisted Driving: The Tesla test was much shorter—about eight minutes—and used an easier route. Zhang described cautious behavior around pedestrians and cyclists, but also reported that the driver had to intervene after the system misread a flashing amber turn signal used in France. That moment points to the challenge of adapting assistance systems to local road rules and signals.
5
7
Xpeng Navigation Guided Pilot: Xpeng’s system reportedly followed speed limits strictly and sometimes drove below them. That contrasts with the reported Momenta drive, in which the vehicle sometimes traveled roughly 10% above the limit. These choices show different approaches to speed and traffic flow; on their own, they do not establish which system is safer or more capable.
5
7
Why the rides do not produce a winner
The tests differed in duration and difficulty: Momenta’s roughly hour-long route included the Arc de Triomphe, while Tesla’s lasted about eight minutes on an easier route. Xpeng’s ride also had its own route and conditions. Without matching routes, conditions and evaluation criteria, these demonstrations cannot fairly rank the systems. A single intervention-free drive—or an intervention during one short ride—also cannot establish overall safety or reliability.
5
7
Progress is not the same as driverless operation
The strongest takeaway is that driver-assistance systems are making visible progress in complex urban settings. Momenta’s drive showed that its system could negotiate a difficult junction in a reported real-world demonstration. The braking and late lane maneuver, the Tesla signal misread and Xpeng’s conservative speed behavior also highlight different challenges: smooth decision-making, accurate interpretation of local rules, and choosing speeds that suit the situation.
These were supervised demonstrations, not evidence that the cars could operate driverlessly. Xpeng’s Paris ride was described as requiring an attentive driver, and the Jaecoo-Momenta challenge was reported as an L2++ drive.
2
31 The demonstrations are best read as useful snapshots of current capability—not a definitive comparison or a guarantee of performance across European roads.