Tesla FSD in Europe: Sweden Approves Public Testing

Tesla is one step closer to conquering the complex roads of Europe. In the latest sign that the continent is finally opening its doors to autonomous driving, the Strängnäs municipality in Sweden has officially granted Tesla permission to begin testing its Full Self-Driving (FSD) software on public roads.

The news was first shared by @LinkN01 on X, who posted a copy of the official decision letter from the municipality’s Technical and Leisure Committee. The approval marks a major win for Tesla, which has been pushing for testing rights in Sweden since early this year to gather critical data in unique Nordic traffic environments.

The Road to Final Approval

While the municipality has given the green light, the testing won’t start quite yet. The permission is valid for one year, but only after Sweden’s Transport Agency (Transportstyrelsen) gives its final stamp of approval. Unlike the unsupervised Robotaxi rides we’ve seen recently in the U.S., these Swedish tests will require a specially trained safety driver behind the wheel at all times.

The approval specifically covers municipal roads in Strängnäs, though Tesla already has permission to test on state roads. According to the decision letter, the municipality believes that automated driving can “significantly improve traffic safety” by reducing human error — the leading cause of most accidents. By collecting real-world data in Sweden, Tesla can better train its neural networks to handle the ice, snow, and low-light conditions common in Northern Europe.

The European Domino Effect

Sweden joining the testing ranks is part of a massive “domino effect” currently sweeping through the European Union. Just a couple of weeks ago, the Netherlands became the first EU country to fully approve FSD (Supervised) for public use, leading Tesla to launch subscriptions and begin its first public rollout of the feature in the region.

Tesla leadership, including Ashok Elluswamy, has previously suggested that Dutch approval would lead the way for the rest of the bloc (and beyond). We are already seeing this bear fruit; Spain has already approved testing, and Italy is reportedly fast-tracking its own approval process. By establishing a presence in Sweden, Tesla can ensure its software is tailored specifically for the diverse driving norms and safety standards of the Nordic market.

What’s Different in Europe?

For Swedish owners hoping to get their hands on the software soon, it is worth noting that the European version of FSD is a slightly different beast than the one in North America. To comply with local regulations, the system includes unique UI changes and strict safety requirements. Owners must also complete a mandatory safety quiz and watch a tutorial video before the feature can be enabled for the first time.

As Tesla continues to refine its FSD v14 and v15 builds, the data gathered from Swedish roads will be instrumental in achieving “superhuman” safety levels globally. With more countries jumping on board every month, the dream of a fully autonomous Europe is looking less like a “someday” project and more like a quickly developing reality.

As Tesla continues to expand the utility of its vehicles, a recent update to the official mobile app has revealed a hidden feature that will soon make towing easier. In app version 4.55.5, Tesla added a new text string pointing to an upcoming capability that will allow owners to test their trailer lights directly from their smartphones.

Solving a Common Towing Headache

Anyone who frequently tows a boat, camper, or utility trailer knows the frustration of the pre-trip light check. To ensure that the trailer’s brake lights, running lights, and turn signals are properly synced with the tow vehicle is a safety and legal requirement.

Historically, this has been a two-person job: one person sits in the driver’s seat pressing the brake pedal and activating the blinkers, while the other stands behind the trailer to confirm everything is illuminated correctly. 

For solo drivers, the task becomes more difficult. The process can involve the driver only checking some features, like turn signals, or more awkward setups, like propping up a phone camera to check that all their trailer lights are working properly.

By integrating a trailer light test directly into the Tesla app, owners would be able to quickly and easily test their trailer lights by standing behind their trailer with their Tesla app.

This feature likely stems from the Tesla Semi, which already includes various testing features for owners to check their truck prior to a trip.

The Code Hint

The upcoming feature was discovered by digging through the source code of the Tesla app version 4.55.5. There was a phrase in the app called MOBILE_APP_FEATURE_TRAILER_LIGHT_TEST , as found by @Tesla_App_iOS.

Tesla frequently updates its mobile app to lay the groundwork for future vehicle features. These backend code strings are reliable indicators of what the engineering team is preparing to launch or is at least testing in the real world.

Coming to the Vehicle UI

While a dedicated trailer light testing mode isn’t available yet in Tow Mode, we imagine that this new string entry means that Tesla will eventually bring the feature to the vehicle UI as well.

With the Cybertruck expanding Tesla’s market into the heavy-duty towing segment, and with the already capable Model X and Model Y, adding quality-of-life towing features is a logical next step for the software team.

Tesla has officially deployed a new machine learning model designed to improve the accuracy of the displayed wait times for Superchargers. In a recent update shared by the official Tesla Charging account on X detailed how it is drastically improving the accuracy of its in-car Trip Planner forecasts by better predicting driver behavior.

The Mixed-Use Challenge

The core problem Tesla engineers faced was accurately predicting wait times at mixed-use locations. Many Supercharger stations are intentionally co-located with busy amenities such as grocery stores, restaurants, and shopping centers.

When driving to a local Supercharger, many drivers don’t add the location to the vehicle’s navigation system, making it harder for Tesla to predict wait times. In order to provide more accurate wait times, Tesla also needs to consider the vehicle’s location and predict whether it believes it’s headed to a nearby Supercharger.

However, this causes another issue. The system can struggle to distinguish between a Tesla pulling into a plaza to plug in and one pulling in just to grab a coffee. This mixed-purpose traffic made predicting the actual charging queue incredibly difficult and often led to inaccurate wait time estimates.

Training the AI on Trajectory Data

To solve this, Tesla developed a new machine learning model specifically designed to predict a driver’s true intent to charge. The model was trained on 9 million miles of aggregated, anonymized vehicle trajectory data collected exclusively within the geofenced areas surrounding global Superchargers.

By analyzing exactly how vehicles physically move, navigate, and park within these specific lots, the system can now accurately distinguish a charging customer from a standard retail shopper based entirely on their trajectory.

The 20% Benchmark

The results of this new model are impressive. According to Tesla, the updated software reduces queue-length estimation errors to about 20 percent.

In real-world terms, this means that even during the busiest holiday travel days, when a station might have a frustrating queue of 10 or more vehicles, the Trip Planner prediction will only be off by a margin of one or two cars. This allows the navigation system to intelligently route drivers to slightly further, but ultimately faster, Supercharger locations to minimize total travel time.

No Update Needed

Since a vehicle fetches Supercharger wait times via Tesla’s APIs, no update is needed to start seeing these changes. Tesla stated that it’s now rolling out this update; however, it’ll likely be limited to certain Supercharger locations as Tesla analyzes the results before rolling it out to everyone.

Preparing for Virtual Queuing

While highly accurate forecasts help drivers plan around congestion, Tesla is also working on a permanent solution to manage the congestion itself. Tesla’s Charging Team is actively piloting a new virtual queue system designed to completely eliminate the physical line at busy Supercharger locations. Tesla has started building out the virtual queue feature in the latest Tesla app.

Instead of navigating a chaotic parking lot and physically waiting in a disorganized line of cars, drivers will soon be able to join a digital waitlist. Whether accessed through the official Tesla app or directly on the vehicle’s display, the system will assign users a spot in the queue and provide real-time updates on their position.

Once a charging stall becomes available, the system will direct the driver to their designated spot.

By combining the new trajectory-based AI forecasting with a structured virtual queue, Tesla is actively ensuring that the public charging experience remains as seamless and stress-free as possible, even during peak holiday travel.