How to increase your Tesla Safety Score

September 30, 2021

By Nuno Cristovao

If you've opted-in to Tesla's assessment of your driving behavior, then you're likely trying to get access to Tesla's FSD Beta. In order to do that you'll need as high of a score as possible. The higher the score, the sooner you'll be granted access to the FSD Beta.

Tesla's very transparent about how your Safety Score is calculated. We'll share some information for each category Tesla is grading and also share some tips on how you can improve your score.

There are five categories that Tesla is actively measuring drivers against. It's important to be familiar with them, as not everything counts toward your Safety Score. Your performance in each category is measured to the median of Tesla's fleet.

Forward Collision Warnings (FCWs)

First, we have Forward Collision Warnings. This is the car giving you a warning that you're going at a high rate of speed relative to the distance you have to the object directly in front of you.

Tesla bases this off the ‘medium' setting in FCW. I'd recommend going to Autopilot > Forward Collision Warning and setting your personal alert level to ‘Early'. This way you'll be made aware of the potential collision before Tesla dings you for it.

There's not much to this one, except to keep your distance and slow down before you get too close to the vehicle or object in front of you.

Hard Braking

This is a tough one and the one I've struggled with the most. You'll want to very gradually slow down when coming to a stop. Quite a bit more than you'd likely think you need to. One harsh stop and you'll be dinged.

Tesla identifies hard braking as a backward acceleration greater than a g-force of 0.3. I personally love to take full advantage of the car's regenerative braking. So if I have a shorter distance to stop, I'll let my foot off the accelerator a little quicker to receive a higher level of regen, but doing so exceeds the 0.3g Tesla specifies and will set you back. Brake early and gradually to avoid any negative scores due to hard braking.

Aggressive Turning

Similar to harsh braking, Tesla is looking at the g-forces here when determining aggressive turning. If the lateral g-force exceeds 0.4g, then your score may be affected.

You'll want to make sure you're not going faster around turns. The tighter the turn, the slower you'll want to go.

Unsafe Following

This is the distance between you and the vehicle in front of you. It's similar to Forward Collision Warnings in that you'll want to make sure you leave enough room to stop between you and the vehicle in front of you, but Unsafe Following is only calculated when you're traveling above 50 MPH. An easy way to avoid penalties here is to use Autopilot with a follow distance greater than 3.

Forced Autopilot Disengagement

This is the act of using Autopilot and having it become disengaged because you didn't heed the warnings of applying pressure to the steering wheel. Not applying enough resistance on the steering wheel will not affect you here, neither will the car beeping. It is only considered a forced disengagement if the vehicle asks you to apply resistance three times and then disables Autopilot for the remainder of the trip.

Keep in mind that Autopilot will also be disabled if you exceed 90 MPH for vehicles with radar or 80 MPH for vision-only vehicles, while on Autopilot.

Tips

Although, I don't personally encourage this tip, as we want everyone to be fair, there is currently a way of having your drive not count toward your Safety Score. If you do a soft reset before you park your car, the drive will not be registered and will not count toward your overall Safety Score. This will likely be fixed in a future update.

Each driving category is weighted differently. So having Autopilot disabled isn't the same as having an instance of harsh braking.

Here is how much each category negatively affects your score, from the worst offender to the least.

Forced Autopilot Disengagement
Hard Braking
Aggressive Turning
Unsafe Following
Forward Collision Warnings

Autopilot

While you're driving on Autopilot your score is not negatively affected, even if the car follows too closely or gets a FCW. Tesla simply ignores any bad driving if Autopilot is engaged. You may want to consider using Autopilot in more situations when you feel it's safe.

If you're using Autopilot on city streets, remember to disengage it with plenty of time before your turn so that you can slow down gradually before taking the turn.

Acceleration

Although acceleration or speeding do not negatively affect your score, it could result in sudden braking or Forward Collision Warnings, which would impact your Safety Score.

Other Metrics

Tesla collects a lot of information on how we use our cars. They use this information to improve features, roll out new ones and make their cars safer.

Although Tesla is showing us these five metrics that they're using to assess your Safety Score, Tesla could potentially use more information than just your Safety Score when choosing the next batch of beta testers. They're probably not looking for individuals who are driving 200 miles a day, or those that are rarely driving. You'll probably want to be in a sweet spot where you drive most days, but aren't taking long trips.

Tesla may also use information that is out of your control such as your geographical region or how long you've had your car.

Use Other Cars

If you're already driving your Tesla most days, don't take the chance of ruining a good score. If you're in a rush or already drove your Tesla today, consider taking a spouse's car when possible.

Save your Tesla for driving on familiar roads, roads where you know when a sharp turn may be coming up or where stop signs are. Use your Tesla when you have the extra time to take it slow and really focus on the road.

Lastly, be safe. We all want a high Safety Score, but at the end of the day we all need to come home in one piece.

If you're looking to figure out how many miles you'll need to drive to reach a specific score, such as 99, I strongly recommend using our Safety Score Calculator that will give you exactly that answer. After adding your data, you can choose a target score and know exactly how much more you'll need to drive to achieve your goal.

New 'Wile E. Coyote' Wall Video Puts Tesla's FSD V13 to the Test

March 21, 2025

By Not a Tesla App Staff

Not a Tesla App

Last week, Mark Ruber, an engineering YouTuber best known for his glitter bombs, released a video where he tested Tesla's Autopilot against various conditions - including the iconic ACME painted wall.

During this test, many people noted that Mark was using Autopilot rather than FSD, even though his video was titled “Can you Fool a Self-Driving Car?”. The Tesla on Autopilot went up against a vehicle equipped with Luminar’s LIDAR rig, running some sort of basic autonomy or safety software.

New Video Tests FSD

Many people were disappointed with Mark’s video and his testing methods, so several creators got to work to actually test out Tesla’s FSD.

Creator Kyle Paul over on X made a much better follow-up video, using both a HW3 Model Y as well as an AI4 Cybertruck. In a relatively unsurprising turn of events, the Cybertruck was successfully able to detect the wall, slowed down, and came to a stop. The Cybertruck was running FSD 13.2.8.

Kyle’s team did a fantastic job building the wall and testing this in a private area using FSD rather than Autopilot. On top of that - they re-tested the results several times and recorded the entire thing in and out. While Mark’s video was more for entertainment, Kyle really set out to prove what would really happen in this unlikely scenario.

Sadly, the HW3 Model Y was unable to detect the wall, and manual intervention was required in each test. While the Model Y was running FSD 12.5.4.2 rather than an FSD V12.6 build, we don’t expect this to have had a significant impact on the test - this is more of an issue with how computer vision analyzes the environment.

There are several major differences between HW3 and HW4. The first is obviously that the version that runs on AI4 is more advanced, as the hardware is capable of processing a lot more data. However, AI4 also features much higher-resolution cameras than HW3, and Tesla recently added the ability for the video feeds to be processed at full resolution on FSD V13. This could have made the difference, although it’s not entirely clear. Perhaps if HW3 gets a version of FSD V13 in the future, HW3 can be retested to see if it passes the “ACME wall” test.

Watch

Kyle’s entire video is below. It’s only 10 minutes long, so definitely give it a watch. Props to Kyle on the quick and thorough execution.

What Does This Mean for FSD?

We broke down Mark’s test - and examined all the little issues that we discovered after doing some in-depth research - you can read our analysis here.

Putting aside the issues with Mark’s testing and instead using the new results - it seems that if you were to have to fight against Wile-E-Coyote and his ACME tools with your Tesla, cartoon logic may win if you’re on an HW3 vehicle. If you’re on an AI4 vehicle, you’ll likely come to a safe stop.

Vehicle depth perception is definitely something that Tesla has been hard at work to improve - and some fairly drastic improvements came with FSD V13 that haven’t been entirely translated to FSD V12 just yet. Future versions of HW3 FSD may be able to determine that the wall is there successfully. So Kyle - if you’re reading this - don’t get rid of that wall. We’d love to see more testing in the future.

However, this entire test scenario is so out of left field… there is a good likelihood this same test would fool some human drivers as well. The most important part is that the future of autonomy will not fall for these tricks, so it's very unlikely for someone to weaponize this idea as it’d only possibly work on a small segment of vehicles.

If Wile-E-Coyote is after you, someone else may drive into the wall before your Tesla does.

Not a Tesla App

We’re not kidding, this really happened already. This isn’t a realistic scenario outside of someone trying to play an insane prank - but it’s good to know that FSD V13 is capable of dodging this.

Tesla Hosts Public All-Hands Meeting for Q1 2025 - Recap & Replay

March 21, 2025

By Not a Tesla App Staff

Not a Tesla App

Tesla regularly holds quarterly all-hands meetings for employees, but last night marks the first time Tesla has live-streamed the event for the public.

The meeting primarily focused on Tesla’s employees, recent achievements, and the future of Tesla. While it didn’t reveal much new information, it was interesting to see Elon Musk candidly engaging with his teams, who seem to genuinely enjoy working with him. Still, there were a few noteworthy takeaways.

As with Tesla’s Earnings Calls and other live events, we’ve put together a concise, easy-to-digest recap of everything discussed.

General Points

Work-related Injuries declined over time
Planning to expand to new markets
Cell Manufacturing
- Continuing to invest in battery supply
- Cheapest, lowest cost per kWh cell
The supercharger network continues to grow
- Vehicle range and charging speed should match the needs of humans and their required time for breaks
- V4 Cabinets, 500 kW for Cybertruck, 1.2MW for Semi
Tesla is building sustainable abundance
- Autonomy is key for this to happen
- Optimus and AI compute
- This appears to be part of Elon’s new mission for Tesla, but it isn’t in the company’s updated mission statement just yet.
Service and Collision Expansion will continue, with a big focus on Service
- Service is what will sell future vehicles
Interesting opportunity to make an electric, supersonic jet
- Elon is stretched too thin but believes Tesla has all the tech in-house to do it
Tesla is in Master Plan Phase III - making all energy on Earth sustainable / abundance for all

Tesla’s Achievements

Sold 7 million vehicles globally
- From just 20 in their first year of production
- 10 million by next year
- 3 million vehicles a year
Giga Berlin has produced 660,000 drive units
Giga Shanghai has built 3m cars
Giga New York has produced 160,000 NACS adapters
GIga Nevada has built 11,000 battery packs in a year
Semi Factory on track for completion
- Over a million, millions possibly
- Will be autonomous in the future
Stationary Battery Storage is a gigantic plus
- Shanghai Megafactory is up and running
- More Megapacks and Powerwalls are being produced than ever
Switchover to New Model Y in a short period was an amazing feat of supply chain engineering

Optimus

Optimus Production Line in Fremont is up and running
- The first Optimus came off the production line this year
- Pilot production this year
- 22 DoF hand and forearm are now in production
Real brain behind Optimus, not on the cloud and not pre-programmed
What’s learned from cars (batteries, electronics, design, manufacturing, FSD) transferred to Optimus
Biggest product of all time
- 5,000 units by the end of the year, 10,000 at max
- 50,000 next year
- Middle of next year for outside of controlled environments
- Will be available late 2026 for employees - who will get first dibs on access to Optimus
Tesla will expand Optimus production to Giga Texas as well

Cybercab and FSD

The Cybercab line will be fully autonomous
- The aim is 5 seconds between producing cars
- 50,000-ton gigacasting machine - make five castings at a time
  - Currently 9,000-ton machines
  - Seems to be an interesting engineering challenge
Progress towards FSD (Unsupervised)
- Cars doing useful work with no one in them (end-of-line to outbound lot)
- Just started at Giga Texas
- FSD will be 10x safer than humans
Autopilot Hardware 4 / AI 4
- Nothing better on the market than AI4 right now
- No point in waiting for AI5 or AI6 as Tesla will continue producing new versions. There will be an AI 7, 8, etc.
Autonomous Teslas will be everywhere
- In 5 years, regulatory approval globally
- Even without Cybercab, there is a gigantic fleet that can be autonomous

Cortex and Dojo

Cortex - AI Training
- One of the most powerful
- 50,000 active GPUs
- 100,000 GPUs coming soon
Dojo now active in Giga New York and Palo Alto
- Now actively working, doing about 5%-10% of the training load