A while back, Morgan Stanley's well-respected automotive analyst Adam Jonas evaluated Tesla (TSLA +0.68%) using a sum-of-the-parts model between artificial intelligence (AI), software, energy, and robotics rather than considering it a traditional automaker. What's interesting is that Jonas believes autonomous driving technology and the robotaxi business drive 41% of Tesla's valuation compared to 34% from its core automotive and energy business and about 25% from Optimus robot potential. So, when robotaxi rival Waymo of Alphabet (GOOG -0.12%)(GOOGL -0.13%) points out why Tesla's driverless technology strategy could have serious drawbacks, investors should take note.
What's going on?
Recently, Alphabet's Waymo co-chief executive officer, Dmitri Dolgov, seemingly took a shot at Tesla when speaking at Y Combinator's Startup School, though he didn't name the automnaker specifically. Dolgov essentially argued that camera-only self-driving technology could be considered "weak sensing" and that the strategy would develop quickly initially before hitting a lower ceiling of capability and performance long term.
Image source: Y Combinator / Waymo co-CEO Dmitri Dolgov at Startup School 2026.
For years, the common argument for a camera-only system was that it's cheaper and that humans rely solely on vision when driving. Therefore, a camera-only system could work adequately for driverless vehicles. Dolgov essentially agreed that a camera-only system could match human performance, but that to build a driverless technology that's safer than humans, which is the entire goal, there needs to be more sensors.
Although Tesla has opted for a camera-only driverless system strategy, which enables the automaker to lower costs, Waymo opts to use three sensor types: cameras, LiDAR, and radar. "These different sensing modalities, they're not backups to each other," Dolgov said during the presentation. The data fuses into a single view of the world that he noted is "vastly superior to what you get with any one sensor."
It's true that using three sensors is better than one unless you believe all three are redundant. In my opinion, they aren't. Consider this simple scenario: A snow storm could cause a whiteout for camera-only systems, which would see next to nothing, while LiDAR in the same scenario would have no problem detecting a human or obstacle on the roadside. Even a fluke event such as mud covering the camera lens could completely shut down the driverless vehicle, whereas a Waymo vehicle with LiDAR and radar could safely navigate back to its home base to clean the camera.
NASDAQ: TSLA
Key Data Points
Falling behind?
For Tesla investors, the criticism about camera-only systems should be concerning because there is truth to it. It's a potential speed bump for Tesla especially when you consider there are other issues with Tesla's driverless technology and robotaxi strategy, such as Tesla being on the hook to replace the self-driving computer in roughly 4 million vehicles, or figure out a way to compensate owners fairly after admitting Hardware 3 isn't powerful enough to deliver the unsupervised self-driving as advertised.
Image source: Tesla.
Another issue for Tesla investors to chew on is that the automaker has yet to deliver much transparency or a timeline for its Cybercab approval process. The vehicle needs approvals to begin charging for rides. Amazon-owned Zoox recently received approval by the National Highway Traffic Safety Administration (NHTSA) to commercially deploy its purpose-built, steering-wheel-free robotaxis, enabling it to officially charge for rides, which it plans to do shortly in Las Vegas.
What it all means
Simply put, investors need to be aware of not only Tesla's camera-only capability for its driverless system but the steps it needs to take for the approval process so that its robotaxi business can truly start expanding. Waymo, among other rivals, has already established a lead in the business compared to Tesla. Considering the latter's valuation is largely believed to be from its robotaxi potential, Tesla needs to play catch up fast.
