BMW, Intel and Mobileye Autonomous Driving For 2021 Conference – Tesla Autopilot Accident Reaction

JUL 3 2016 BY MARK KANE 32

Press conference BMW Group, Intel and Mobileye (from left to right): Brian Krzanich, CEO of Intel, Harald Krüger, Chairman of the Board of Management of BMW AG, and Amnon Shashua, Mobileye Co-Founder, Chairman and CTO Professor.

Press conference BMW Group, Intel and Mobileye (from left to right): Brian Krzanich, CEO of Intel, Harald Krüger, Chairman of the Board of Management of BMW AG, and Amnon Shashua, Mobileye Co-Founder, Chairman and CTO Professor.

BMW Group @ CES 2016, BMW i Vision Future Interaction

BMW Group @ CES 2016, BMW i Vision Future Interaction

BMW has announced alliance with Intel and Mobileye to bring fully autonomous cars to streets by 2021.

The foundation of the fully autonomous cars for BMW will be electric BMW iNEXT.  The trio intends to create an open platform for autonomous driving, from door locks to the data center.

The open platform means also that other car makers could use the system or maybe even join the project.

The first prototype will be demonstrated soon, while next year BMW plans tests in fleets.

“The common platform will address level 3 to level 5 automated driving and will be made available to multiple car vendors and other industries who could benefit from autonomous machines and deep machine learning.

The companies have agreed to a set of deliverables and milestones to deliver fully autonomous cars based on a common reference architecture. Near term, the companies will demonstrate an autonomous test drive with a highly automated driving (HAD) prototype. In 2017 the platform will extend to fleets with extended autonomous test drives.”

And yes, the conference was held July 1st, so the recent new that the NHTSA was investigating a Tesla Model S fatal accident while in Autopilot mode put a fairly dark mood on the event, and some of the conference questions surrounded that event.

Scene Of May 7th Tesla Model S Fatality (via ABC News/Bobby Vankavelaar)

Scene Of May 7th Tesla Model S Fatality (via ABC News/Bobby Vankavelaar)

Possibly as a result, BMW’s Harald Krüger often noted that “safety comes first” during the event, while  adding the death was “really very sad” and that he felt “technologies today are not ready for serious production”.

Tesla’s statement on the accident (which the Florida Highway Patrol indicated was caused by a tractor trailer that “failed to yield right-of-way” when driving across a highway perpendicular to the Model S) did recognize a potential short-coming of the Autopilot system in avoiding a collision in a special chain of circumstances.

“Neither Autopilot nor the driver noticed the white side of the tractor trailer against a brightly lit sky, so the brake was not applied. The high ride height of the trailer combined with its positioning across the road and the extremely rare circumstances of the impact caused the Model S to pass under the trailer, with the bottom of the trailer impacting the windshield of the Model S.”

The aftermath of the accident doesn’t put the focus on whether or not the Tesla was at fault (as by the rules of the road it wasn’t), but rather if autonomous driving needs to be set to a higher standard.

Should predictive analysis be greater?  Should greater diagnostic abilities of on-road conditions be in place? 

For example:  better recognizing the side of a trailer vs an overhead road sign, or awareness of a collateral circumstance that could lead to an at-risk situation – such as reckless/improper driving which is not occurring in the current path of a vehicle, but could impact it in the future.  

These are things an alert human mind still digests more fully and quickly than autonomous drive technology today.

EE Times also reported on Amnon Shashua, Mobileye’s co-founder, CTO and chairman from the event on the Tesla accident:

“Companies need to be very transparent about limitations of the system. It’s not enough to tell the drivers to be alert, it needs to tell them why they need to be alert. It’s not just a lawyer talk.”  He added, “I think a company as reputable as BMW does things right in that respect.”

On timeline/functionality for autonomy:

Amnon Shashua (Mobileye) said that Level 3 (eyes-off) in 2021 would be limited to highways, but that on a highway “you are completely safe”

“That means you can really take eyes off, and there is a significant grace period from the time when the system is compromised until you really need to take control. And if you don’t take control, the system will know how to stop aside slowly in a safely manner.”

Level 5 autonomy (driver off) is also in the cards for 2021, but still with some limits – specific locations and routes.

“We are envisioning fully autonomous driving in a geo-fenced area. Take a city like Munich. A certain (geo-fenced) area provides safe, completely Level-5 autonomous driving.”

BMW Vision Next 100 showcases autonomous driving

BMW Vision Next 100 showcases autonomous driving

Press release:

“Munich – July 1, 2016. BMW Group, Intel, and Mobileye are joining forces to make self-driving vehicles and future mobility concepts become a reality. The three leaders from the automotive, technology and computer vision and machine learning industries are collaborating to bring solutions for highly and fully automated driving into series production by 2021.

The future of automated driving promises to change lives and societies for the better. But the path to get to a fully autonomous world is complex and will require end-to-end solutions that integrate intelligence across the network, from door locks to the data center. Transportation providers of the future must harness rapidly evolving technologies, collaborate with totally new partners, and prepare for disruptive opportunities.

Steady. Steady. Use the car for balance if you must BMW CEO Harold Kreuger

Steady. Steady. Use the car for balance if you must BMW CEO Harold Kreuger

Together with Intel and Mobileye, the BMW Group will develop the necessary solutions and innovative systems for highly and fully automated driving to bring these technologies into series production by 2021. The BMW iNEXT model will be the foundation for BMW Group’s autonomous driving strategy and set the basis for fleets of fully autonomous vehicles, not only on highways but also in urban environments for the purpose of automated ridesharing solutions.

BMW Group, Intel and Mobileye are convinced that automated driving technologies will make travel safer and easier. The goal of the collaboration is to develop future-proofed solutions that enable the drivers to not only take their hands off the steering wheel, but reach the so called “eyes off” (level 3) and ultimately the “mind off” (level 4) level transforming the driver’s in-car time into leisure or work time. This level of autonomy would enable the vehicle, on a technical level, to achieve the final stage of traveling “driver off” (level 5) without a human driver inside. This establishes the opportunity for self-driving fleets by 2021 and lays the foundation for entirely new business models in a connected, mobile world.

On July 1, 2016, the three partners were present at the BMW Group Headquarters in Munich to express their commitment to strive for an industry standard and define an open platform for autonomous driving. The common platform will address level 3 to level 5 automated driving and will be made available to multiple car vendors and other industries who could benefit from autonomous machines and deep machine learning.

The companies have agreed to a set of deliverables and milestones to deliver fully autonomous cars based on a common reference architecture. Near term, the companies will demonstrate an autonomous test drive with a highly automated driving (HAD) prototype. In 2017 the platform will extend to fleets with extended autonomous test drives.

“Today marks an important milestone for the automotive industry as we enter a world of new mobility. Together with BMW Group and Intel, Mobileye is laying the groundwork for the technology of future mobility that enables fully autonomous driving to become a reality within the next few years,” said Mobileye Co-Founder, Chairman and CTO Professor Amnon Shashua.

“Mobileye is proud to contribute our expertise in sensing, localization, and driver policy to enable fully autonomous driving in this cooperation. The processing of sensing, like our capabilities to understand the driving scene through a single camera already, will be deployed on Mobileye’s latest system-on-chip, the EyeQ®5, and the collaborative development of fusion algorithms will be deployed on Intel computing platforms. In addition, Mobileye Road Experience Management (REM) technology will provide real-time precise localization and model the driving scene to essentially support fully autonomous driving.”

Intel brings a comprehensive portfolio of technology to power and connect billions of smart and connected devices, including cars. To handle the complex workloads required for autonomous cars in urban environments Intel provides the compute power that scales from Intel® Atom™ to Intel® Xeon™ processors delivering up to a total of 100 teraflops of power efficient performance without having to rewrite code.

“Highly autonomous cars and everything they connect to will require powerful and reliable electronic brains to make them smart enough to navigate traffic and avoid accidents,” said Intel CEO Brian Krzanich. “This partnership between BMW Group, Intel and Mobileye will help us to quickly deliver on our vision to reinvent the driving experience. We bring a broad set of in-vehicle and cloud computing, connectivity, safety and security, and machine-learning assets to this collaboration enabling a truly end to end solution.”

With its Strategy Number ONE > NEXT, the BMW Group has developed its framework to remain the driving force behind premium individual mobility. This approach will become driving reality with the BMW iNEXT model in 2021, heralding a new era of mobility.

“At the BMW Group we always strive for technological leadership. This partnership underscores our Strategy Number ONE > NEXT to shape the individual mobility of the future,” stated Harald Krüger, Chairman of the Board of Management of BMW AG. “Following our investment in high definition live map technology at HERE, the combined expertise of Intel, Mobileye and the BMW Group will deliver the next core building block to bring fully automated driving technology to the street. We have already showcased such groundbreaking solutions in our VISION NEXT 100 vehicle concepts. With this technological leap forward, we are offering our customers a whole new level of sheer driving pleasure whilst pioneering new concepts for premium mobility.”

EE Times, Hat tip to George K!

Categories: BMW

Tags: , , ,

Leave a Reply

32 Comments on "BMW, Intel and Mobileye Autonomous Driving For 2021 Conference – Tesla Autopilot Accident Reaction"

newest oldest most voted

BMW RE: Tesla AP “not ready for serious production”.

BMW and others are about to get a free ride on Tesla being the first to tackle an issue we’ll all still be facing in 2021. By that time there will still be shortcomings that will kill people, but where the most progress will be will not be in tech. It will be in customers accepting we can’t completely trust these systems. For that, they should be thanking Tesla.

Once again, the ones with the most resources fail to pioneer.

Apple vs Blackberry?

Sorry, but the pioneer is not always in the best position to see the bigger picture. That’s reality.

The last problem Tesla has is “big picture”. Funding the time gap to that picture. That may be their problem.

If automation gets shelved because of Tesla, it’ll be bittersweet. Believe me. I think there are a number of people reading this website who’d make driving illegal, and they need to be slowed down. Tesla already assumes we all have the same driving skills, and that already takes some of the polish off for me. Let’s get back to electrification?

pjwood1 said:

“Tesla already assumes we all have the same driving skills…”

I confess I don’t understand what you’re saying here. Would you care to expand on that?

Seems to me..

The quickest (inevitable?) way forward in autonomous drive..

1. Require a “networked transponder” in every vehicle
2. Restrict/phase out human drivers

Not that I’m advocating the above, but I do wonder if that’s the direction all this will go, and how fast will it go there?

“Phase out human drivers”!

Skynet woudn’t say better!

It is time to stop foolishness.

… and just what “foolishness” would that be?

You remove human control without consent, you open the door for system malfunction deaths, you allow hackers a way to instantly mass killing millions, you allow people at electronics controls to selectively organize a kill which leads to Orwellian power. That kind of foolishness. Do you need more?

Nah, that’s enough.

.. add,

I’m gonna recommend you abstain from riding on the airlines.

Airplane have a completely different situation. At first take off and landing is usually done manually by the pilots, second there are almost no obstacles in the air and if there are the plane has a serious radar system as well as assistance from traffic control. Plus there are two pilots on board and most people aren’t able to fly themselves even less a Boeing 777 or an Airbus A380.
In space you could not steer yourself since trajectories must be computed to be very precise. You also can’t pilot because there is no real reference to do it, although the Apollo 13 crew did a famous manual steering and Neil Armstrong landed manually after problems with autopilot.
I will trust a self driving car when there is a true AI on board and one that passed his driving license. I will also require that he has no backup somewhere and that his circuits also die in case of a crash to be sure he does care about not crashing.

Paranoid much? Human drivers kill people and must be eliminated….

Start with asking why an autopilot allowed the car to speed more than 20 mph above the speed limit in the first place.

Anything that requires a 100% replacement of existing infrastructure is dead, full stop.

Autonomous cars will have to live with non-autonomous traffic, as well as other environmental hazards (trees across the road, etc).

Right. I can easily envision a near-future in which most cars will be both wirelessly connected and more-or-less fully autonomous. I can easily imagine that all major urban centers will have “nodes” which control traffic, monitoring flow and redirecting traffic around accidents and other obstacles, and perhaps playing “umpire” in assigning right-of-way at intersections and when cars need to change lanes in dense traffic. Perhaps these nodes will operate, and be placed, like cell towers are now. And just as with cell towers, there will be areas too remote and too sparsely settled to justify installation of such nodes. So yeah, self-driving cars are gonna have to be able to deal with operating independently, relying only on GPS data and the car’s own sensors. The will also have to be equipped to deal with vehicles driven by humans, for some decades to come. There would be too much public pressure against banning human-driven cars from the roads, until nearly all cars without autonomous capability have disappeared from the roads because they’re too old. I don’t see a universal ban on human-driven vehicles ever happening. Off-road vehicles, for example, will continue to allow human drivers, because in that area, the fun… Read more »

And in that nice future there will be now and then a virus killing people not from the flue but because suddenly all cars will start to go into head on collisions at fast speed or start to drive on the pedestrians or crash in trees or drive off cliff. They will even have names like the memorable 2078 wall virus that made 37809 cars collide in walls at 150 mph.
Great future indeed. We may even end up having more road deaths than less because of those unexpected events.

I just wonder what’s wrong with a good driver, perhaps using automatic breaking and the like, but otherwise driving responsibly. Wreckless drivers must be taken off the road not the good ones.

When it comes to automotive innovation BMW is ready for nothing, but another press conference.

Tesla has already done more for car safety than all other car makers combined.

Not ready… And how you determine readiness? Tesla autopilot is the only tool today that can cut through bullshit and provide hard data about what really happened. And 35000 victims die every year, with nearly 250 due to this precise underside collision.

A test between the Mercedes e-class autopilot and the Tesla autopilot showed already that Tesla is lagging in some important fields. So take the fanboy glasses down, Tesla often let the customer doing their beta test drives.

This was 2 years ago. Your glasses are stained with oil.

LOL, in that test Tesla won. Tesla has one every other test comparing self driving features. Sounds like you need to put on some glasses.

A test between the Mercedes e-class autopilot and the Tesla autopilot showed already that Tesla is lagging in some important fields.

In a field marked by rapid development, isn’t that exactly what any informed person would expect? Presumably the Mercedes e-class “autopilot” is also lagging in some important fields, compared to Tesla’s semi-autonomous driver assist features.

More to the point, no autonomous (or semi-autonomous) system in any production car today deserves the term “autopilot”. Perhaps the suite of hardware and software controlling Google’s fleet of self-driving cars is advanced enough to deserve that term.

BMW doesn’t even have an electric sedan yet and they already want to force feed the selfdriving foolishness on us. No thanks.
It is time to let go on that bad idea just like battery exchange for cars was.
Lets concentrate on more energy in the battery and megawatt power at superchargers that is what’s important.

We want electric cars, not autonomous cars. Instead of expending resources on a nonexistent problem let’s focus on getting the damn ICEs off the road.

I personally share this sentiment, but I’m not among the world’s many people who spend hours in traffic every week. 🙂

“Possibly as a result, BMW’s Harald Krüger often noted that “safety comes first” during the event, while adding the death was “really very sad” and that he felt “technologies today are not ready for serious production”.”
Sounds like BS to me. A C&D comparison from a few months ago showed that such a system was already in the 750i. Automakers like to talk a lot as if they aren’t putting that system yet into production, while Tesla is wreckless for doing so, when the truth is they already have them in production.

“Amnon Shashua (Mobileye) said that Level 3 (eyes-off) in 2021 would be limited to highways, but that on a highway “you are completely safe””
What happens if you are on a highway that allows left turns, like the one where the accident happened? Probably in Europe such roads don’t exist, but in the US it does.

In Europe “Highway” is not a network but a class of roads with the following properties :
– motor vehicles only with minimum performance requirements (no pedestrians, no underpowered vehicles)
– dual carriage way separated by a barrier
– emergency shoulder all the way (including bridges and tunnels)
– fences to prevent pedestrians/wildlife from crossing the road
– no crossing EVER (only high speed entrance and exits, roads cross on separate grades)

I find it amazing the the USA, the country of the car, does not have a word for it and still calls “highway” roads that are clearly unsuitable for heavy traffic.

Isn’t just about every automotive manufacturer who wants in on autonomous drive using mobileye technology ? … including Tesla?
(Tesla, GM, Nissan, Ford, Volvo, VW, BMW, Honda .. at least, that I’ve seen reference to partnerships)

My understanding (without re-googling the whole thing) is that Tesla uses mobileye but adds a significant “self learning (AI) layer to it.

/typical corporate crap, … everyone’s going to poo poo it until they’ve got their product (mostly from the same supplier) ready to launch, .. then it’s the greatest thing ever

“For now, a thriftier alternative to Lidar are high-tech cameras. Mobileye, an Israel-based company with a market cap of around $10 billion, is emerging as the dominant supplier here. More than 90 percent of carmakers have partnered with Mobileye, the company said. Tesla buys its cameras, which are laced with Mobileye’s advanced software and chips. Earlier this month, GM said it was working with Mobileye to test self-driving features on the hybrid Chevy Volt.

Mobileye’s full hardware and software package, which includes a 360-degree view around cars, can cost less than $1,000 for car companies, said chief communication officer and SVP Yonah Lloyd. “For the car industry, cost is a major consideration,” he added.

Cameras are not only cheaper, but they often have better resolution, and can see certain elements on the roadways — lane markers, traffic lights — that Lidar misses. A few startups, like Nauto and Cruise, are deploying cameras coupled with smart computer-vision algorithms to retrofit cars with driver-assistance technology that set the stage for full autonomy.”

Meet the Companies Building Self-Driving Cars for Google and Tesla (And Maybe Apple)
http://www.recode.net/2015/10/27/11620026/meet-the-companies-building-self-driving-cars-for-google-and-tesla

Thanks very much for your most informative post, carcus! I had gotten the idea that Tesla wasn’t using its car cameras for autonomous driving, because their cars also have front-facing radars and ultrasonic movement sensors. Clearly I was wrong. If Tesla is eschewing lidar because of mere cost, that’s… at the least very unfortunate, if not completely wrong-headed. Software imaging using a steroscopic camera system is only as good as the software programmed to recognize objects. Speaking as a computer programmer, object recognition software is woefully inadequate for safe control of a moving vehicle. Even a slow-moving vehicle, let alone one moving at high speed! If software can more readily identify lane markings using camera images than with lidar, then by all means let autonomous cars use cameras for that. But for detection of obstacles, moving or otherwise — especially other vehicles which are moving — I doubt anyone who was really informed on the limitations of software to recognize images from a camera would ever trust his life to a vehicle controlled by that kind of software. Tesla’s comment about the recent fatality being a result of the car’s semi-autonomous system being unable to distinguish between the side of… Read more »

The cost of the scanning Lidar is about the same as the cost of a P90D, just for the sensor.

That’s even more crazy, not only do we have selfdriving foolishness but they are already controlled by a monopoly. We are even closer to big brother than we could think off. The NSA will soon be able to send Merckel’s car in a lake if it suits them instead of just listen to her phone. Edward Snowden is never going to take a car anymore.

Has InsideEVs ever done an article detailing the different “levels” of driving autonomy? If so, I missed it.

Here’s what appears to be a good article with succinct summaries describing the US Department of Transportation’s National Highway Traffic Safety Administration (NHTSA) defined five different levels of autonomous driving:

http://www.techrepublic.com/article/autonomous-driving-levels-0-to-5-understanding-the-differences/