In profitability push Mobileye dumps LiDAR, slashes workforce
- Reference: 1725905803
- News link: https://www.theregister.co.uk/2024/09/09/mobileye_cut_lidar_layoff/
- Source link:
The Israeli self-driving tech player today [1]announced plans to end the internal development of next-generation LiDAR tech, eliminating around 100 jobs in the process. Mobileye cited a variety of reasons for the decision including "LiDAR is less essential to our roadmap for eyes-off" self-driving systems.
Instead, Mobileye said it's going to focus on the "substantial progress" made on its EyeQ6-based computer vision perception system, the "increased clarity on the performance of our internally developed imaging radar," and ever-more affordable third-party LiDAR kit to fill the gaps.
[2]
The decision to stop work on LiDAR systems "has no bearing on Mobileye's commitment to the development of our in-house imaging radar, which … is expected to enter production next year," the team said. "In terms of Mobileye's internal sensor development, imaging radar is a strategic priority."
[3]
[4]
LiDAR has looked like a long shot in the self-driving world for a while, with the firms pushing its development becoming some of [5]the biggest losers in the self-driving industry in recent years.
Reasons given for why avoiding LiDAR have been varied, with Tesla boss Elon Musk, long an opponent of the tech, calling it "stupid, expensive and unnecessary" in 2019. Former Tesla AI chief Andrej Karpathy has also expressed doubts about LiDAR's practicality as a piece of self-driving technology due to its need to store and update maps, and the expensive cost of hardware.
[6]
In short, this could simply be Mobileye making a business adjustment to pursue a more promising technology.
Then again, things aren't going great for Mobileye
Seeing Mobileye's decision to scrap its LiDAR R&D unit as indicative of the future of self-driving cars may be a bit hasty: The outfit isn't doing that well.
Mobileye's most recent earnings report was disastrous, with the operation [7]slashing its revenue and profit forecasts due to volatile demand, causing shares to plummet earlier this month.
Axing the LiDAR division and its 100-person team isn't Mobileye's first 2024 cut - it also [8]eliminated its aftermarket solutions unit in March, shedding 130 jobs dedicated to retrofitting driver assistance systems to older vehicles.
[9]Self-driving cars safer in sunlight, twilight another story
[10]Chinese car brands hit accelerator on road tests for level three autonomous driving tech
[11]GhostStripe attack haunts self-driving cars by making them ignore road signs
[12]Two big computer vision papers boost prospect of safer self-driving vehicles
This was followed by the news that [13]struggling Intel , which [14]acquired Mobileye in 2017 for $15 billion, was [15]considering selling as much as 88 percent of its Mobileye stake in a bid to drum up funds and save cash.
Mobileye hasn't turned a profit in the past couple of years and 2024 is looking like its third consecutive losing year. News of its recent struggles has pushed shares down to levels below what the company was worth when Intel [16]took it public in 2022.
Whether the decision to eliminate its LiDAR unit is a strategy shift is unclear and we haven't heard back from Mobileye. But we do notice that it made one thing perfectly clear in its press release: The move will save lots of cash.
[17]
"Operating expenses for the [LiDAR] R&D unit are expected to total approximately $60 million in 2024," the outfit said. "While this action is not expected to have a material impact on Mobileye's results in 2024, it will result in the avoidance of [LiDAR] development spending in the future." ®
Get our [18]Tech Resources
[1] https://www.mobileye.com/news/mobileye-to-end-internal-lidar-development/
[2] https://pubads.g.doubleclick.net/gampad/jump?co=1&iu=/6978/reg_onprem/systems&sz=300x50%7C300x100%7C300x250%7C300x251%7C300x252%7C300x600%7C300x601&tile=2&c=2Zt9v_yonb2P5fVKwFPeLxgAAAc0&t=ct%3Dns%26unitnum%3D2%26raptor%3Dcondor%26pos%3Dtop%26test%3D0
[3] https://pubads.g.doubleclick.net/gampad/jump?co=1&iu=/6978/reg_onprem/systems&sz=300x50%7C300x100%7C300x250%7C300x251%7C300x252%7C300x600%7C300x601&tile=4&c=44Zt9v_yonb2P5fVKwFPeLxgAAAc0&t=ct%3Dns%26unitnum%3D4%26raptor%3Dfalcon%26pos%3Dmid%26test%3D0
[4] https://pubads.g.doubleclick.net/gampad/jump?co=1&iu=/6978/reg_onprem/systems&sz=300x50%7C300x100%7C300x250%7C300x251%7C300x252%7C300x600%7C300x601&tile=3&c=33Zt9v_yonb2P5fVKwFPeLxgAAAc0&t=ct%3Dns%26unitnum%3D3%26raptor%3Deagle%26pos%3Dmid%26test%3D0
[5] https://www.theregister.com/2022/10/18/selfdriving_tech_values/
[6] https://pubads.g.doubleclick.net/gampad/jump?co=1&iu=/6978/reg_onprem/systems&sz=300x50%7C300x100%7C300x250%7C300x251%7C300x252%7C300x600%7C300x601&tile=4&c=44Zt9v_yonb2P5fVKwFPeLxgAAAc0&t=ct%3Dns%26unitnum%3D4%26raptor%3Dfalcon%26pos%3Dmid%26test%3D0
[7] https://www.reuters.com/technology/mobileye-cuts-annual-revenue-forecast-shares-slump-2024-08-01/
[8] https://en.globes.co.il/en/article-mobileye-to-close-aftermarket-unit-lay-off-130-1001473991
[9] https://www.theregister.com/2024/06/19/autonomous_driving_dim_conditions/
[10] https://www.theregister.com/2024/06/04/china_autonomous_driving_tests/
[11] https://www.theregister.com/2024/05/10/baidu_apollo_hack/
[12] https://www.theregister.com/2024/05/29/computer_vision_nature_self_driving/
[13] https://www.theregister.com/2024/09/06/intel_foundry_in_jeopardy/
[14] https://www.theregister.com/2022/10/03/intel_mobileye_ipo_fabs/
[15] https://www.bloomberg.com/news/articles/2024-09-05/intel-is-said-to-explore-sale-of-part-of-stake-in-mobileye
[16] https://www.theregister.com/2022/10/18/intel_mobileye_ipo_16b_valuation/
[17] https://pubads.g.doubleclick.net/gampad/jump?co=1&iu=/6978/reg_onprem/systems&sz=300x50%7C300x100%7C300x250%7C300x251%7C300x252%7C300x600%7C300x601&tile=3&c=33Zt9v_yonb2P5fVKwFPeLxgAAAc0&t=ct%3Dns%26unitnum%3D3%26raptor%3Deagle%26pos%3Dmid%26test%3D0
[18] https://whitepapers.theregister.com/
I thought the whole point of LIDAR was it provided a proper 3D view of the world without the ambiguity involved in deciphering a stereo image of ill-defined objects?
I guess radar has the advantage of seeing through fog (though not smart to be speeding along because you can see, when others can't see you) and maybe less likely to suffer white-out in direct sun, though often you won't get the same degree of spatial resolution in tolerably small packages. I can't see it being much cheaper though.
Also I don't get this aspect of needing well-defined maps - cars need to deal with changes in the road that are not documented due to road works, accidents, trees down, etc, so anything that needs a map beyond basic route planning sounds crazy to me.
Hmm. In *theory* LIDAR is the way to go
Resolution of human sight.
Potentially very small "antennae" IE lenses.
Maybe go the other way and go sonar instead, given the actual useful time windows* needed?
*It's not about physical range, it's about response time. Worst case 150mph -->240Kmh-->67m each way. --> ~2.5 scans/sec with sonar.
Let's see if I have this straight. Tesla -- which apparently can't build a lane-keeping/collision avoidance system that isn't perpetually just one or two more fixes away from working right doesn't like LIDAR. Waymo on the other hand -- which many people think to be the clear leader in the field -- uses LIDAR (albeit in conjunction with just about every other remote sensor technology known to man). Which would you make you feel safer if you saw it bearing down on you while you were crossing the street -- a Tesla or a Waymo Jaguar I-PACE?
Perception
crash.