Tesla recently started rolling out an update that enabled the cabin radar in recent Model Ys and added support for first-row cabin sensing. This functionality not only lets Tesla replace the seat occupancy sensors in the front seats but also gives them a better idea of the size of the passengers — critical information when an accident occurs.
However, Tesla has now revealed that it won’t just be recent versions of Model Y that will receive these updates. Other recent vehicles also have a dormant in-cabin radar that will be activated in a future update.
Child Presence Detection
This front-row occupant sensing is only half of the update. Tesla also plans to add support to the rear seats that will let it detect occupants left in the vehicle. In a post on X, Tesla stated that at least 37 children die per year from heatstroke, just in the United States, according to the NHTSA. Tesla’s cabin radar will help reduce the number of deaths by detecting children left in the vehicle and alerting the owner or turning on the HVAC system.
The new radar has several safety-focused features, and they’re all designed to help protect someone left behind in a vehicle. First up - the radar will detect and classify passenger sizes to determine which airbags to deploy dynamically, helping to reduce injuries from shorter individuals sitting up front.
Additionally, the radar can differentiate an adult from a child or baby - and automatically and dynamically make the necessary safety adjustments on the fly. Tesla will also be able to detect heart rate and breathing through the radar - and has a series of safety steps it can take from here.
If a child or baby is detected being left behind in the vehicle, the Tesla app will immediately notify you and turn on the HVAC system to ensure their safety. In the case of an emergency, the vehicle will automatically call emergency services and report its location.
As a reminder, while Tesla vehicles outside of the United States support eCall, vehicles within the US cannot call 911 unless a phone is connected to the vehicle via Bluetooth. We hope that Tesla looks to add similar capabilities in North America to improve safety capabilities.
Supported Models
Although the cabin radar has only been activated in Model Ys produced in the last couple of years, Tesla said in their post on X that the radar is also available in recent Model 3s and the Cybertruck.
We did a little digging and confirmed that the part numbers for the cabin radar are the same for the 2024+ Model Y, 2024+ (new) Model 3, and the Cybertruck. So, these vehicles should also receive the latest cabin radar features.
Although the 2023 Model Y features a different part number, 2022 and 2023 Model Ys have also been receiving the front-row cabin sensing feature in Tesla’s 2025.2.6 update, but it is less clear whether they’ll receive the rear seats update later this year. The Model S and Model X also include a cabin radar, but the part number also differs from the cabin radar for the 2024 Model 3 and Model Y.
While the new Model Y isn’t available to check against on Tesla’s Electronic Parts Catalog, it looks like Tesla has been shipping the new 4D cabin radar in at least the 2024+ versions of their Model 3, Model Y and Cybertruck.
We expect rear-seat sensing to arrive on vehicles with the latest cabin radar later this year, while it’s unclear whether the other vehicles with the older cabin radar will receive it.
Tesla mentioned that rear-seat support will arrive in Q3 2025 for the refreshed Model Y. Based on the different interior sizes and spaces, sensor calibration and testing for other models may come slightly later.
Tesla’s Optimus humanoid robot is back in the spotlight, and this time it's showing off some impressive new footwork. In a pair of videos shared on X this week, Optimus demonstrated a surprising level of agility and coordination while dancing. The first demonstration had a precautionary safety cable, but the second demo was impressive without support.
Tesla has been making some astounding advances in Optimus’ unique FSD model, particularly through simulation and reinforcement learning. Optimus has come a long way since tripping over its own feet.
This first glimpse showed Optimus performing a simple series of dance-like movements. While there was a safety cable visible, indicating the relatively early nature of the test, the fluidity of the movements was noteworthy. Check out those feet—maintaining balance on two legs may be easy for humans, but it’s much harder for a robot as it shifts its center of gravity.
Milan Kovac, a member of the Optimus AI team, provided some additional context on this initial video. He mentioned that more was coming shortly, and the team has been hard at work. What’s key here is that he emphasized that the routine was entirely trained in simulation with reinforcement learning. He also mentioned that there are many optimizations and fixes already put in place for the sim-to-real training code. Finally, regarding the cable, he mentioned it's there in case of a fall and that it is not actually holding Optimus up directly.
Dance Baby, Dance
True to Kovac’s promise, the Tesla Optimus X account followed up with a second, more impressive video, declaring that it was just “getting warmed up.”
This time, Optimus engaged in a far more dynamic EDM-style dance, and crucially, without a safety cable or restraint. Optimus put on a good show of its balance, quicker movements, and an even greater range of dynamic motion, all untethered.
Murtaza Dalal, another Optimus AI team member, commented on this cable-free performance, stating, "As promised, the team is crazy fast :)" He pointed to the core methodology enabling such rapid progress: "Sim2real RL is the key to getting next level agile, dynamic motions. It’s also the key to precision and robustness."
Sim-to-Real Learning
The recurring theme from the Optimus team is the power of sim-to-real transfer using reinforcement learning. This approach involves training the AI model extensively in a simulated virtual environment, where it can learn complex behaviors (like walking, balancing, and now dancing) through trial and error at an accelerated pace. The best part is that it can all be done without risking damage to the physical hardware and done across multiple nodes of Tesla hardware. Imagine thousands of Optimus bots learning to dance all at once - except virtually.
Once the AI masters these skills in the simulation, the challenge lies in transferring that learning effectively to the real-world robot - which is the sim-to-real step.
These latest dance demonstrations suggest Tesla is making some fairly substantial strides in bridging that gap, allowing Optimus to translate simulated learning into real-world physical competence.
Rapid Development
While dancing may seem like a novelty, it serves as a compelling visual demonstration of Optimus’ advanced capabilities in balance, coordination, and dynamic movement - all essential for performing useful tasks in real-world environments. This progress builds on previous milestones, like Optimus learning to walk.
The "crazy fast" development pace lauded by Dalal indicates that Tesla is treating Optimus with the same iterative intensity it applies to its vehicles and FSD software. Each new demonstration, from sorting objects to walking and now dancing, offers a glimpse into a future where humanoid robots could play an important role in manufacturing, logistics, and eventually daily life.
The ability to quickly iterate and improve in simulation and then successfully deploy those improvements to the physical robot is crucial. While Optimus might not be ready for household chores just yet, its new dance moves are a clear sign that FSD is learning and evolving at an impressive rate.
A recent sighting at Giga Texas by drone operator and detail hound Joe Tegtmeyer has really caught our eye. A red Model Y was driving within the factory premises with its front and rear sections heavily camouflaged in Tesla’s standard black coverings. To top it off, there wasn’t just one, but two vehicles.
Tesla usually conceals vehicles when it has something to hide or test, and this timing suggests to us that a new variant of the world’s best-selling SUV is on its way. This leads us to two theories - a refreshed Model Y Performance or the anticipated, more affordable E41 Model Y.
Caught something interesting today … two heavily camouflaged cars driving on the W side of Giga Texas. Several viewers noted this in my short video on the autonomous drive.
— Joe Tegtmeyer 🚀 🤠🛸😎 (@JoeTegtmeyer) May 14, 2025
The front fascia, headlights, and hood are almost entirely obscured, with similar coverings applied to the rear, hiding the taillight design and bumper. This is a classic practice to keep new design elements under wraps before an official unveiling, and this is exactly what we saw with the Model 3 Highland and Model Y Juniper projects. The rest of the vehicle, including the central body and roofline, appear fairly consistent with the current Model Y.
Refreshed Model Y Performance?
One compelling possibility is that Tesla is preparing for the launch of the refreshed Model Y Performance. This speculation gains traction when we look at the launch of the refreshed Model 3 Performance. The Performance variant launched several months after the Long Range AWD and RWD variants and came with a bespoke front fascia. That fascia included air ducts for improved aerodynamics and brake cooling, helping to differentiate it both visually and by performance from other refreshed Model 3s.
It stands to reason that Tesla is preparing to launch the refreshed Model Y Performance. The extensive covering on the front fascia could be hiding a more aggressive, aerodynamically optimized bumper and air ducts to match the Model 3 Performance.
More Affordable Model Y?
Alternatively, this mystery vehicle could be the prototype of the previously discussed affordable variant of the Model Y, known as E41. Tesla has been working on further cost reductions for its highest volume vehicle, which includes cost-cutting components and features, as well as working on manufacturing efficiencies.
Tesla is focusing on making its existing models, like the Model Y, more affordable. In this case, the camouflage might be concealing simplified exterior features designed to help reduce production costs. This could include a revised front-end design, along with different headlights or taillight assemblies. Tesla could even forgo the distinctive front lightbar, as well as the afterburner-style rear lightbar, if it contributes to worthwhile savings.
This would be a step beyond the recently introduced Model Y Long Range RWD, which improves the vehicle’s range and lower the price by $4,000.
For now, the identity of this camouflaged Model Y remains a mystery. The fact that it is out and about at Giga Texas suggests that whatever changes Tesla has incorporated are significant enough to warrant covering up the front and rear of the vehicle.