Autopilot sure has come a very long way since the “early days”. It’s amazing how much has been accomplished in such a short time. Nevertheless, back when Tesla wasn’t making it’s own Autopilot processors, the technology already was extremely advanced and way ahead of it’s time. So, how does it compete with the AP2 version that preceded it? Was AP2 even better than AP1?
The second Autopilot (AP2) version struggled initially with the new Nvidia processor replacing the Mobileye processor. With later updates, the AP2 version steadily grew to a more advanced and reliable version of Tesla’s Autopilot. Tesla had to build the complete system from the ground up to support their vertical .
If you’d like to see all the initial AP1 and AP2 differences and how all this came to be, then check out the rest of this article. I explain how to upgrade, some issues that they faced (and still do) and some of the technical stuff as well. It was a fun topic to research. Check it out!
Understanding Tesla’s New Autopilot System
Tesla launched their self-driving feature, dubbed the Tesla Autopilot, back in 2014. The cutting-edge driver assistance technology known as Autopilot improves both security and comfort while behind the wheel.
The Autopilot system is a big departure from the earlier models, which relied on drivers to keep their hands on the wheel at all times. Tesla’s new Autopilot system relies on sensors and cameras to keep the car in control. Each new Tesla comes with a total of eight cameras mounted on the outside as well as advanced image processing software.
One of the biggest changes with this new Autopilot system is that it can handle more complicated situations. For example, it will be able to parallel park without help from the driver. Another change is that the car will be able to drive in low light conditions, something that was difficult for earlier models of Teslas.
Pro Tip: How can you let a Tesla parallel park on its own?
One would usually have to turn to the “P” on the stock when parked next to an available parking spot. Now if you see the “P” appear when parked next to a parking spot, all you can do is put it into reverse where it will alert you to go into auto-park. From there you can let the car do its thing. Enjoy easy parking from now on out!
If you use the Autopilot feature appropriately, you’ll cut down drastically on the effort you put in while driving.
However, this feature is not available in older models, so you won’t find an older car retrofitted with the Autopilot system. Instead of using radar, the newer camera-based Tesla Vision system is now being integrated into all cars manufactured for the North American market.
The Main Differences Between Tesla AP1 and Tesla AP2
Aside from being the latest version of the Autopilot feature, there are many important differences between AP2 and AP1.
The base version referred to as Autopilot 1 is actually not designed by Tesla! A software company named Mobileye created the system in partnership with Tesla. It was implemented as an assistive feature in cars produced between 2014 and 2016.
The AP1 hardware consists of a single camera attached to the front, paired with a radar (with a range of 525 feet (160 meters) connected to Mobileye’s computing software. There are also 12 sensors fitted around the car with a range of 16 feet (5 meters).
On a somewhat related note, I wrote an article comparing Tesla’s Enhanced Autopilot (EAP) vs. Full Self Driving (FSD). Check it out to find out what works best for you!
Just because it’s old doesn’t mean the Tesla AP1 doesn’t pack a punch. It covers some of the main requirements that drivers need: adaptive cruise control and assistive features such as lane switching and centering. Since many drivers only use their adaptive driving software on the freeway, having an AP1 doesn’t necessarily make your car les capable.
The main features available with AP1 are:
- Cruise Control
- Lane changing
- Automatic parking
- Car Summoning
Keep in mind that auto-steering is available on freeways, and lane changes should be initiated by the driver.
Tesla’s second version of the Autopilot came out in 2016. It featured a lot more bells and whistles than its older counterpart. AP2 combines the radar with several cameras placed around the car, giving it a complete 360-degree view.
An inbuilt system replaced Mobileye’s processor with an NVIDIA AI computer. This processor was initially not on par with Mobileye’s one, but now it is up to speed, thanks to several updates.
The AP2 hardware consists of 3 front cameras instead of 1, 4 side cameras (2 on each side), and 1 rear-facing camera. The front radar (range 525 feet (160 meters)) is now accompanied by 12 ultrasonic sensors (range 26 feet (8 meters)). This helps the NVIDIA processing platform make real-time adjustments to the car’s driving capabilities.
The newer version of the Autopilot has a few more tricks up its sleeve than its predecessor. Drivers can now make use of the following features:
- TACC: Traffic-Aware Cruise Control
- Auto steering on freeways
- Automatic lane changing (still needs to be initiated by the driver)
- Automatic Parking
- Summoning of a parked car
Did You Know
Did you know:
The Tesla Model X, equipped with AP2, is rated as the safest vehicle by the National Highway Traffic Safety Administration.
The hardware of the AP1 only consists of a single camera, but the AP2 includes eight cameras placed all around the vehicle to provide a complete 360-degree view. This makes it possible for the vehicle to be much more aware of its surroundings.
For instance, the system will be able to determine if the side lane contains a car or a bike. In addition, radar and ultrasonic sensors have been upgraded to detect objects at greater distances.
The original AP1 was created with a processor from Mobileye, which relies on complex code to identify the many things it encounters. Tesla collaborated with Mobileye to further improve it for its cars, but the technology could only be advanced to a certain point.
In contrast, the NVIDIA DRIVE PX 2 is a robust graphics processing unit (GPU) developed specifically for use in machine learning and neural networks.
The Tesla Model S and Model X produced from October 2014 to October 2016 come fitted with the Autopilot 1 hardware. Tesla models S and X produced after 2016 come fitted with the AP2 hardware.
How To Determine Whether Your Tesla has AP1 or AP2
This can easily be done by looking at the exterior of the car. Older cars with AP1 have no visible cameras on the sides, whereas the newer AP2-fitted vehicles do have them.
Take a look at the side panel indicators on your Tesla. If you can see camera holes in them, then it’s fitted with AP2. If you see nothing aside from a standard Tesla turn signal/indicator, you have AP1 or a car with no Autopilot features at all.
You can also look at the front of your car—specifically, the rearview mirror portion of your windshield. If you see a camera there but no cameras on the sides of the vehicle, then it’s an AP1 car. If you don’t see any cameras, then the car isn’t fitted with Autopilot hardware.
If you want to check it through the vehicle’s touchscreen, simply open the software tab in your vehicle settings, and click on “additional vehicle information.” Here, you’ll be able to check out the exact version of Autopilot that your car is fitted with.
Upgrading AP1 to AP2
If you want your car to not be street legal and become a disaster waiting to happen, then sure, you can try to upgrade your AP1 to AP2. But Tesla doesn’t offer such services. It would be achievable if you were willing to spend an extravagant amount of money on a 3rd party modifier (since Tesla wouldn’t go through the trouble of such a thing).
Honestly, it would be much cheaper to just trade in your old Tesla for a new one. The money and effort spent in retrofitting your Tesla is just not worth the effort. A person did do it, and you can check the article here on retrofitting your Tesla.
In the middle of July 2017, Tesla announced the release of AP2.5, a minor upgrade to the AP2 hardware package. It had an improved front radar, a Continental Advanced Radar Sensor ARS410, which had a slightly more extensive range (558 feet), as well as the NVIDIA DRIVE PX 2 with secondary node-enabled functionality.
It was a small change, and there were no obvious practical advantages associated with it at this time.
|Tesla AP1||Tesla AP2/AP2.5||Tesla AP3|
|Automatic Lane Change||Not Possible||Feature Included||Feature Included|
|Smart Summon Car||Not Possible||Feature Included||Feature Included|
|Traffic Sign Recognition||Feature Included||Not Possible||Not Possible|
|Auto Steering||Not Possible||Not Possible||Feature Included|
Full-Self Driving: Tesla’s Latest Addition to the AP Series
Late in April 2019, the FSD Computer, also known as Hardware 3 or AP3, started deploying in all Teslas. In 2019, upgrades were made available to AP2 users who had previously bought FSD. FSD stands for Full Self Driving Computer, the latest in Tesla intelligent driving tech.
Current AP2 owners have the option to update their NVIDIA graphics cards. This upgrade will be provided at no additional cost to customers who selected Full Self-Driving but will be an optional purchase for those who did not. In April 2019, Tesla started installing its Full Self-Driving computer in all of its cars.
Here’s a video that breaks down the difference between AP3 and AP2:
Random Fun Fact:
FSD cars are not perfect. In fact, they make mistakes just like humans do. However, because they are able to learn from their mistakes, full self-driving cars are much more successful than traditional cars when it comes to safety.
Autopilot vs. Full Self Driving
Autopilot is a type of driver assistance system that allows a car to drive itself under certain conditions. It can help avoid collisions, make turns, and change lanes.
Full self-driving, on the other hand, is a much more advanced system that allows a car to drive without any help from the driver. This would include driving in all conditions, including heavy traffic and stop-and-go traffic. However, full self-driving is still in development and is not available on all cars.
The Autopilot consists of several features, which are:
- Lane Assist
- Collision Avoidance Assist
- Speed Assistance
- Auto High Beam
- Traffic-Aware Cruise Control
The Full Self Driving addition has all the features mentioned above, as well as the following features:
- Navigation on Autopilot
- Automatic Lane Changes
- Automatic parking
- Summoning a car from a parked position
If you flash your high beams it will bring up a menu where you can toggle your high beams (this menu is like the wiper speeds settings menu). From there you can select your light settings for Off, Parking, On, or Auto.
Upgrading From AP2 to FSD?
AP2 cars can be upgraded to FSD through your Tesla account, as long as they are equipped with the necessary hardware. If you’re a bit worried about whether it’s worth the hassle, then no worries!
Tesla offers you the opportunity to try out their Autopilot/Full Self Driving capabilities at their Tesla store locations before buying it. For more information, you can check out Tesla’s website.
Tesla’s Autopilot Isn’t Flawless
If you’re trusting a machine with your life, it’s on you if you suffer because of it. Tesla emphasizes the requirement that a driver needs to be completely aware and in control of the car. Tesla’s self-driving and assistive capabilities are not meant to replace the driver.
In an interview on CBS, Elon Musk himself emphasized that these systems are “far from perfect.” Several factors can reduce the effectiveness of your Tesla’s Autopilot, such as the weather, bright lighting, obstruction of the cameras, and extremely narrow and winding roads.
Tesla’s website instructs all drivers to keep all cameras and sensors clean and free from obstructions to fully exploit the system.
The Future of Self-Driving Technology
While we’re still a long way off from entirely autonomous cars, Tesla (and other car manufacturers) are working tirelessly to make that concept a reality as soon as possible. For now, remember to judge self-driving capabilities on the following factors:
- Limitations: Where is the system unable to work correctly? Today, many systems are still only effective on freeways and not as much on city streets.
- Accuracy: How accurate is the system in calculating when to change speed, and how long can it manage to be semi-autonomous without any driver input?
- Emergency actions: This is perhaps the most challenging thing for Autopilot systems to get right, given so many unexpected things may happen in real life. This is why most systems demand the driver to keep aware at all times.
- Smooth transitions in speed: while cruise control has been around for a while now, it pays when the system can successfully maintain a set speed and change speed smoothly.
- Human error checking: How well does the system ensure that you are attentive? Good systems should issue a warning if they sense that the driver is inattentive. This is usually done by checking that both hands are on the wheel and whether the driver is looking at the road. For this purpose, newer Teslas have cameras that note eye movement.
In 2021, Tesla recorded an accident for every 4.59 miles (7.39 kilometers) driven by a vehicle that had Autopilot enabled. Cars driving without Autopilot recorded an accident every 2.42 miles (3.89 kilometers).
Frequently Asked Questions
How do I calibrate my Autopilot system?
To calibrate your Autopilot system, the cameras must finish a process called “self-calibration.” Calibration is done after driving 20 to 25 miles (32 to 40 km), depending on the road and environment.
Can I sleep while the car’s on Autopilot?
You can’t sleep while your Tesla is on Autopilot. You can trust the Autopilot to assist you in driving, but you must remain alert and in control of the vehicle at all times.
If you found this article helpful, feel free to give us a thumbs up by sharing it on your social media platform of choice. You can also check out these 15 great gift ideas for yourself or a Tesla fanboy.
Contact Us if you have any questions or queries.