Intel Is Giving Up On Its AI-Powered RealSense Cameras

In a statement to CRN, Intel said it was "winding down" RealSense and transferring the talent and computer vision tech to efforts that "better support" its core chip businesses. Engadget reports: Questions surfaced about the fate of RealSense after the team's leader, Sagi Ben Moshe, said he was leaving Intel two weeks ago. RealSense aimed to make computer vision more flexible and accessible. A company or researcher could buy cameras to aid everything from robot navigation through to facial recognition, and there was even a developer-focused phone. It was never a truly mainstream product, though, and ASI VP Kent Tibbils told CRN that there were few customers buying RealSense cameras in any significant quantities. It wasn't really a money-making division, even if the work helped Intel's other teams.

For Intel, there's likely a simpler answer: it wants to cut ballast. CEO Pat Gelsinger wants Intel to reclaim the chipmaking crown, and that means concentrating its resources on design and manufacturing capabilities. No matter how successful RealSense is, it's a potential distraction from Intel's latest strategy.

Intel has too high of cost to pursue this field.

[jonsmirl]

Here is a much more cost effective option.
Firefly CAM-C1126S2U [www.firefly.store]

Re:

[Kisai]

The Realsense camera is as close as you get to Apple's ARKit for the PC, and unfortunately, like pretty much everything Intel makes that is not a CPU, Intel tends to drop the ball and either produces product that is redundant with something better that already exists or doesn't produce them in the quantity and price to justify continued software investment.

I have a realsense camera, I've been able to do very little with it, and I can buy a used iPhone X/XS off ebay for less.

Intel is turning into Google

[Junta]

So many of their 'expansion' projects getting canned left and right.

Classic Intel move!

[Gravis Zero]

1) Massively fund project in a competitive emerging field.
2) Corner market with typical anti-competitive practices and sell a product at a loss.
3) Decide it isn't profitable enough and drop the project entirely, leaving a desolated marketplace in their wake.

Fuck you, Intel. Go bankrupt.

Re:

[Tablizer]

Per #2, there is no significant market. That's why they are getting out. Well, I suppose cornering a market of say 37 people with 34 sales is technically still cornering it. "I'm King of the outhouse!"

Re:

[Tablizer]

> but why not try to create a new market?

I'm sure they thought they were innovative. Maybe they were, but it still was a financial flop. The Xerox Star GUI work-station was indeed innovative in its day. But it cost two arms and a leg. Steve Jobs figured out how to slice one arm off in his copy.

Buying

[JBMcB]

I was looking at buying a new webcam a couple of years ago and looked into the RealSense camera. I couldn't for the life of me figure out how to actually buy one. There was an developer's edition that cost something like $400. There were links to third party resellers who didn't have them. There was tons of information for developers, and nothing indicating how you would actually go about buying one as a consumer.

So, that might be one of the problems. Microsoft marketed their Kinect camera for PC better than Intel marketed their RealSense, and the Kinect wasn't even a PC product.

Re:

[jpapon]

I use tons of Realsense cameras at work... they're not hard to buy. You go on Intel's website and you order them.

Re:

[Ostracus]

Few more "tons" and they might have stayed open.

Re:

[tlhIngan]

I use tons of Realsense cameras at work... they're not hard to buy. You go on Intel's website and you order them.

The OP was looking for an off the shelf device a consumer can buy I believe. The one you buy are developer kits, which is how Intel packages the technology.

You're supposed to then integrate the technology into your product.

So it technically wasn't a product for purchase.

Webs

[JBMcB]

That's how it works now. A few years ago it was a lot harder to track them down. I think it was right at the edge of them releasing the second version, whatever that was.

Sounds like Qualcomm

[Snotnose]

Invest heavily into Globalstar, digital theater, Mirasol displays, Media-flo, and I don't remember what all. I spent years at Qualcomm, great company to work for, these were some biggies while I was there.. Whenever they ventured outside of their core baseband/radio chips they crashed and burned.

another product to the dust bin

[Thijson]

Spent 10 years at the company. Some of the other boondoggles, WiMax, Itanium, wearables (NTG), digital home (DHG), McAfee purchase, Windriver purchase, LCOS. Billions spent with very little to show for it. In some cases, the market segment didn't go away, it's just that Intel couldn't compete in it.

and wireless sensors

[rapjr]

Intel sunk a lot into research in the wireless sensor field with the idea that they were going to make the SoC MCU/radio/sensor and all that sensor data would also drive purchases of servers. Just as there started to be early markets they abandoned it saying they didn't thing they could make _enough_ money. I think Intel has no idea how to deal with truly competitive markets because they have been subsidized by the USA in the name of national defense for so long.