I just came back from driving about 200 miles (350 km) using Tesla’s (Supervised) Full Self-Driving, and the system is getting better, but it’s also getting more dangerous as it gets better.
The risk of complacency is scary.
Last weekend, I went on a road trip that covered about 200 miles from Shawinigan to Quebec City and back, and I used Tesla’s (Supervised) Full Self-Driving (FSD), v12.5.4.1 to be precise, on almost the entire trip.
Here’s the good and the bad, and the fact that the former is melting into the latter.
The Good
The system is increasingly starting to feel more natural. The way it handles merging, lane changes, and intersections feels less robotic and more like a human driver.
The new camera-based driver monitoring system is a massive upgrade from the steering wheel torque sensor that Tesla has used for years. I only had one issue with it where it kept giving me alerts to pay attention to the road even though I was doing just that, and it eventually shut FSD down for the drive because of it.
But this happened only once in the few weeks since I’ve used the latest update.
For the first time, I can get good chunks of city driving without any intervention or disengagement. It’s still far from perfect, but there’s a notable improvement.
It stopped to let pedestrians cross the street, it handled roundabouts fairly well, and it drives at more natural speeds on country roads (most of the time).
The system is getting good to the point that it can induce some dangerous complacency. More on that later.
As I have been saying for years, if Tesla was developing this technology in a vacuum and not selling it to the public as “about to become unsupervised self-driving”, most people would be impressed by it.
The Bad
Over those ~200 miles, I had five disengagements, including a few that were getting truly dangerous. It was seemingly about to run a red light once and a stop another time.
I say seemingly because it is getting hard to tell sometimes due to FSD often approaching intersections with stops and red traffic lights more aggressively.
It used to drive closer to how I’ve been driving my EVs forever, which consists of slowly decelerating using regenerative braking when approaching a stop. But this latest FSD update often maintains a higher speed, getting into those intersections and brakes more aggressively, often using mechanical brakes.
This is a strange behavior that I don’t like, but I started at least getting the feeling of it, which makes me somewhat confident that FSD would blow that red light and stop sign on those two occasions.
Another disengagement appeared to be due to sun glare in the front cameras. I am getting more of that this time of year as I drive more often during the sunsets, which happen earlier in the day.
It appears to be a real problem with Tesla’s current FSD configuration.
On top of the disengagement, I had an incalculable number of interventions. Interventions are when the driver has to input a command, but it’s not enough to disengage FSD. That’s mainly due to the fact that I keep having to activate my turn signal to tell the system to go back into the right lane after passing.
FSD only goes back into the right lane after passing if there’s a car coming close behind you in the left lane.
I’ve shared this finding on X, and I was disappointed by the response I got. I suspected that this could be due to American drivers being an important part of the training data, and no offense as this is an issue everywhere, but American drivers tend not to respect the guidelines (and law in some places) of the left lane being only for passing on average.
I feel like this could be an easy fix or at the very least, an option to add to the system for those who want to be good drivers even when FSD is active.
I also had an intervention where I had to press the accelerator pedal to tell FSD to turn left on a flashing green light, which it was hesitating to do as I was holding up traffic behind me.
Electrek’s Take
The scariest part for me is that FSD is getting good. If I take someone with no experience with FSD and take them on a short 10-15 mile drive, there’s a good chance I get no intervention, and they come out really impressed.
It is the same with a regular Tesla driver who consistently gets good FSD experiences.
This can build complacency with the drivers and result in paying less attention.
Fortunately, the new driver monitoring system can greatly help with that since it tracks driver attention, unlike Tesla’s previous system. However, it only takes a second of not paying attention to get into an accident, and the system allows you that second of inattention.
Furthermore, the system is getting so good at handling intersections that even if you are paying attention, you might end up blowing through a red light or stop sign, as I have mentioned above. You might feel confident that FSD is going to stop, but with its more aggressive approach to the intersection, you let it go even though it doesn’t start braking as soon as you would like it to, and then before you know it, it doesn’t brake at all.
There’s a four-way stop near my place on the south shore of Montreal that I’ve driven through many times with FSD without issue and yet, FSD v12.5.4 was seemingly about to blow right past it the other day.
Again, it’s possible that it was just braking late, but it was way too late for me to feel comfortable.
Also, while it is getting better, and better at a more noticeable pace lately, the crowdsource data, which is the only data available as Tesla refuses to release any, points to FSD being still years away from being capable of unsupervised self-driving:
Tesla would need about a 1,000x improvement in miles between disengagement.
In fact, the crowdsource data shows a regression on that front between v12.3 and v12.5.
I fear that Elon Musk’s attitude and repeated claim that FSD is incredible, combined with the fact that it actually getting better and his minions are raving about it, could lead to dangerous complacency.
Let’s be honest. Accidents with FSD are inevitable, but I think Tesla could do more to reduce the risk – mainly by being more realistic about what it is accomplishing here.
It is developing a really impressive vision-based ADAS system, but it is nowhere near on the verge of becoming unsupervised self-driving.
FTC: We use income earning auto affiliate links.More.
Mercedes-Benz says its upcoming entry-level CLA will set the bar for EV range, charging speed, and efficiency with advanced new tech. Ahead of its arrival, Mercedes is giving us a closer look at the sporty electric car.
“With the CLA, we are ushering in a new era at Mercedes-Benz,” CEO Ola Källenius said after driving the new EV at its Immendingen site for the first time.
The new model will be the first of a new family of vehicles, kicking off “a new chapter for the entry-level segment at Mercedes-Benz.” Mercedes said the new electric CLA “is designed to set new standards in range, efficiency and charging speed.”
Based on Mercedes’ next-gen MMA platform, the CLA integrates tech from its ultra-efficient Vision EQXX concept and brings it into mass production.
Mercedes said it developed a new electric drive unit in-house to ensure high performance and efficiency.
Källenius showed off the “secret” CLA EV during the event, revealing new features and details, including the new headlight design.
Mercedes previews new entry-level CLA EV
With the headlights being the first thing you see at night, Källenius said the design had to be perfect. One of the coolest design details is the Mercedes star logo integrated into the headlights. Although it looks pretty, Källenius said the design is also about safety and not blinding others on the street.
After hopping into the new EV, you can hear Källenius talk to the vehicle, and Mercedes’ virtual assistant responds. It’s “like you have a butler in your car,” Mercedes CEO explained.
The new entry-level model will be loaded with Mercedes’ latest software and connectivity tech. It will be the first to be equipped with the new MB.OS, or “the brain” behind its next-gen vehicles, as Mercedes calls it.
During the test drive, the new model shows off its sporty design and dynamic driving as it rips around the test track. “This is not even the AMG version,” Källenius said.
With a two-speed transmission, “we can perfectly calibrate at every speed that you are driving how the electric motor works” to maximize efficiency, Källenius explained.
Mercedes leader said the CLA EV will have “incredibly fast” charging. With a new high-powered charging system, “charging almost becomes like fueling.”
Mercedes will officially unveil the new entry-level CLA EV next year, with deliveries expected in 2026. The new model will rival Tesla’s Model 3 and BMW i4, with expected prices around $50,000. It will also be available in a hybrid version.
FTC: We use income earning auto affiliate links.More.
Unplugged Performance has unveiled its UP.FIT Tesla Model S Plaid police patrol car at Sema – becoming the world’s quickest police patrol car.
The Tesla tuner, through its UP.FIT division, has unveiled several new police vehicles based on Tesla vehicles, but this one is the first to be based on Tesla’s top performance vehicle: Model S Plaid.
This groundbreaking vehicle is designed to meet the high-performance demands of first responders, showcasing the future of electric police vehicles – albeit being a bit of an overkill with a 0-60 mph acceleration in 2 seconds.
Built on the Tesla Model S Plaid platform, the UP.FIT Plaid Pursuit brings a host of modifications and advanced technologies aimed at enhancing high-speed pursuits and traffic patrol operations. The vehicle will be delivered to the Los Angeles County Sheriff’s Department Motorsports community outreach program, underscoring a pivotal shift towards electric vehicles in public service roles.
The UP.FIT Plaid Pursuit model incorporates a proprietary electrical wiring harness designed specifically for UP.FIT Tesla vehicles, allowing for the integration of custom forward- and rear-facing warning lights into the front and rear glass.
According to UP.FIT, this enhances visibility compared to traditional rooftop light bars. Additionally, side skirt lighting and a push-bumper equipped with further warning lights and integrated high- and low-frequency emergency sirens ensure that this vehicle is fully compliant with Title 13 regulations in California.
To complement its exterior modifications, the UP.FIT Plaid Pursuit features performance upgrades from Unplugged Performance, including enhanced braking components and UP Forged lightweight wheels. These enhancements not only boost durability but also provide superior handling in pursuit scenarios. UP leveraged its experience building performance kits for performance-oriented Tesla owners.
This electric patrol car is not only duty-ready but also represents a zero-emission alternative to conventional patrol vehicles. With rapid acceleration and an impressive range of 345 miles, the UP.FIT Plaid Pursuit offers law enforcement agencies a powerful and sustainable option for their operations.
Ben Schaffer, CEO of Unplugged Performance, commented
“The UP.FIT Plaid Pursuit embodies everything we value in electric police vehicles. From outstanding acceleration and range to reduced maintenance and fuel costs, this platform equips law enforcement with the latest tools to enhance their mission while saving taxpayers’ money and promoting a cleaner future.”
The UP.FIT Plaid Pursuit will be displayed at the SEMA FutureTech Studio this week.
Electrek’s Take
This looks incredible. However, I think the Model S Plaid is a bit of an overkill. The Model S Long Range is $15,000 cheaper and the difference is an extra second 0-60 mph and 130 mph top speed rather than 200 mph.
A 3-sec 0-60 mph will smoke most cars, and while the top speed could technically be useful in high-speed pursuits, many police departments abandon pursuits when they reach high speeds because it is deemed too dangerous.
I’m not saying there are never any pursuits above 130 mph, but it is fairly rare.
UP’s Model Y police patrol vehicle is also an cheaper option that reduces the total cost of ownership when accounting for gas savings.
FTC: We use income earning auto affiliate links.More.
XPeng Motors celebrated its annual Tech Day event in China earlier today, which was renamed “AI Day” for good reason. The topic of artificial intelligence dominated the Chinese mobility company’s incoming technologies, which were abundant. We got updates on a new in-house AI computing chip, a new Xpeng platform to support Level 4 autonomous robotaxis, flying cars, and eVTOLs, a new humanoid robot called “Iron,” complete with an existential launch video, and a surprising adoption of combustion-powered extenders that deliver up to 970 miles of additional range.
Today marked XPeng Motor’s sixth consecutive Tech Day event – something we look forward to each fall as it is sure to be filled with exciting updates and debuts from a mobility company that continues to expand and dabble in a little bit of everything, not just passenger EVs.
Last year, we saw the debut of XPeng’s X9 MPV, a humanoid robot called the PX5, and plenty of AI-centric software upgrades. Since then, we’ve seen the X9 officially launch in China along with a rollout of OTA updates to XPeng’s operating systems, including its XNGP ADAS.
This past April, during the Beijing Auto Show, XPeng gave the public a preview of some of the technologies explained further during today’s Tech Day event. That included AI-centric chips, Level 4 autonomous robotaxis plans, and a new neural network called “XNet.” We also saw some fresh updates, including a new humanoid robot, an eVTOL, and a new gas range extender that could deliver some of the farthest-driving vehicles in the world. Let’s dig in.
Source: XPeng Motors
XPeng shares flying car update during Tech Day 2024
There’s a lot to unfold here, but we’ll start with a mainstay in XPeng news, flying cars through its air mobility entity, AeroHT. Today’s talks were led by the “Land Aircraft Carrier” – an eVTOL/vehicle combo that debuted last year.
As we previously reported, the revolutionary land and air vehicle is slotted for scaled production in 2026, and AeroHT recently broke ground on the facility in China that will build them. There is not much of an update here. Still, XPeng founder and chairman He Xiaopeng reiterated that the Land Air Carrier will showcase its first public flight demonstration later this month before pre-orders open in December.
In addition to its modular Land Air Carrier and eventual plans for a bonafide flying BEV with propellers on its roof, Xiaopeng shared that AeroHT is also developing a dedicated full tilt-rotor eVTOL featuring six seats and hybrid power. The aerial vessel will have a maximum range of 500+ km (311 miles) and reach a top speed of 360 km/h (224 mph).
These flying cars and eVTOLs will be powered by a new AI “Turing” chip developed in-house by XPeng.
Source: XPeng Motors/AeroHT
In-house AI chips and a new platform
XPeng renamed its Tech Day event “AI Day” this year for a reason. Artificial intelligence was at the forefront of most of He Xiaopeng’s conversations on stage, and the debut of the mobility company’s new Turing chip was also a part of the event.
The company showcased a new Turing AI Intelligent Driving System, powered by XPeng’s proprietary Turing AI chip, described as “a world-first chip designed for AI vehicles, robots, and flying cars.” The new chip features a 40-core processor and supports models with up to 30 billion parameters, delivering the power of three high-performance chips in one.
Xiaopeng explained that the Turin chip has already completed over 2,700 functional verifications in a mere 40 days and has achieved three times the industry standard for development efficiency during that time.
Source: XPeng Motors
In addition to powering XPeng’s flying cars and humanoid robots (more on that below), Xiaopeng used the 2024 Tech Day event to unveil a new Canghai Platform that will enable full Level 4 autonomous driving. XPeng has already been promising fully autonomous robotaxis by 2025, and we now have a better idea of the technology that will support that, again, centered around its new Turing AI chip.
The company explained that the new platform would act as the previously teased neural network for XPeng’s new line of AI-centric vehicles, supporting driving (with or without a human in the driver’s seat) with enhanced safety features, 33x bandwidth, and 12x faster camera image processing, thus creating “a foundation for full-scenario AI-enabled driving experiences.”
The “Hawkeye” pure vision ADAS system can see the world around it at 720 degrees without blind spots. Per XPeng during Tech Day:
XPeng’s Turing AI system marks a pivotal step toward L4 autonomy, with continual model updates from cloud-based models, ensuring a highly responsive, ever-evolving smart driving experience.
In addition to autonomous robotaxis, XPeng’s Turing AI chip will power a new humanoid robot called “Iron” which made its official debut at Tech Day.
Source: XPeng Motors
XPeng introduces a new humanoid robot called “Iron”
One of the stars of today’s event was Iron, a serious upgrade to the robots XPeng has debuted at Tech Days in the past. With today’s debut, XPeng becomes the latest tech company developing and implementing impressive robot technology that could one day take over full assembly of its vehicles.
Per XPeng, the Turing AI chip enables the Iron bots to think and remember like humans (spooky) and also enables its hands and feet to move autonomously. The robot’s hands also use a 1:1 human hand size, with 15 degrees of freedom.
The debut included an interesting video in which Iron steps into a room and admires famous sculptures like Rodin’s “The Thinker” and “Discobolus” by Myron. Iron even copies the Discobolus pose, as seen in XPeng’s video below:
XPeng’s Tech Day event includes a new gas range extender
Perhaps the biggest surprise from XPeng’s 2024 Tech Day was the news that it is shifting its business strategy from all-BEV to include some combustion with a new gas-powered range extender.
On the stage in Guangzhou earlier today, He Xiaopeng unveiled XPeng’s Kunpeng Super Electric System. Xiaopeng says this next-generation solution “brings together XPeng’s advancements in latest AI-defined mobility innovations to deliver an unparalleled charging and range experience.”
The system consists of an 800V platform powered by XPeng’s battery technology, which supports 5C fast charging and can charge from 10% to 80% in 12 minutes. New to the party is a hybrid silicon carbide coaxial electric drive, which, according to XPeng, will offer industry-leading efficiency and reliability. The range extender is also extra quiet, outputting noise levels as low as 1 dB.
The automaker has also once again integrated AI into the power optimization system, which enables a seamless transition between pure electric and range-extended driving modes. As you can see from the Tech Day image above, XPeng’s new system offers 430 km (267 miles) of all-electric range and a combined range of over 1,400 km (870 miles).
That’s a lot to take in, but it’s all for now. Check back with Electrek soon as we await the official launch of XPeng’s next BEV model, the P7+. In the meantime, you can watch the full 2024 Tech Day presentation on XPeng’s Weibo page.
FTC: We use income earning auto affiliate links.More.