3 ms·
You also don't want AR glasses. The battery alone is a hassle, add the heating/weight of the device (same issues as VR) and a pinch of motion sickness and the f
by ly3xqhl8g9 3y ago
You also don't want AR glasses. The battery alone is a hassle, add the heating/weight of the device (same issues as VR) and a pinch of motion sickness and the faulty input haptics and you are back to the trusty ol' laptop, if you want to actually get something done, or to the smartphone if you are just consuming content. It's just a bad idea all around, perhaps it's easier to hack the optic nerve and feed images into the brain that way.
There is nothing AR glasses can achieve at an industrial site (their main use currently with HoloLens) that a decent digital twin, being able to quickly zoom in/out across hundreds of yards of wiring/machineries, doesn't obtain easier/faster/better/safer/more comfortable. Just what the doctor ordered for a tired technician, crawl under the machinery to check for the AR gimmick some dude in a $4,000 chair designed, only for the thing to be completely unable to map the AR model to the physical object because there's barely any light under a 25 tonne industrial whatever.
- mschuster91 3y agoAR has its uses though. In the military, for example... imagine a squad of soldiers, with every soldier being digitally linked to a central command station that knows the precise location of each unit. A soldier could then see on their glasses - even at night - where their unit mates are, or get commands and status updates from leadership instead of by (often times barely understandable) radio.
- ly3xqhl8g9 3y agoIf they can stop themselves from vomiting, maybe [1] [2]. I am more afraid of drone swarms with pytorch object detection obliterating everything in sight [3], or what Geoffrey Hinton said about the self-healing mine field [4]. Was reading now the "HoloLens environment considerations" [5], most funny aspect: "Wormholes. If you have two areas or regions that look the same, the tracker may think they're the same. This results in the device tricking itself into thinking it's somewhere else." [1] October 2022, "Microsoft Hololens causes US army nausea and headaches", https://www.avinteractive.com/markets/government-and-defence/microsoft-hololens-causes-us-army-nausea-and-headaches-18-10-2022/ https://www.avinteractive.com/markets/government-and-defence... [2] January 2023, "Microsoft has laid off entire teams behind Virtual, Mixed Reality, and HoloLens", https://www.windowscentral.com/microsoft/microsoft-has-laid-off-entire-teams-behind-virtual-mixed-reality-and-hololens https://www.windowscentral.com/microsoft/microsoft-has-laid-... [3] https://fedscoop.com/darpa-drone-swarms-with-1000-unmanned-aircraft-within-5-years https://fedscoop.com/darpa-drone-swarms-with-1000-unmanned-a... [4] https://www.cs.toronto.edu/~hinton/minefield.pdf https://www.cs.toronto.edu/~hinton/minefield.pdf [5] https://learn.microsoft.com/en-us/hololens/hololens-environment-considerations#wormholes https://learn.microsoft.com/en-us/hololens/hololens-environm...
- mschuster91 3y ago> what Geoffrey Hinton said about the self-healing mine field Thankfully, that is more-or-less unfeasible in reality... hard to imagine such a minefield actually working out. The mines would need to be capable of autonomous movement in difficult terrain, would require at least a month worth of battery capacity and recharging capability, and it would be easy to jam their c&c frequency to completely kill that "self healing" capability.
- ly3xqhl8g9 3y agoSure, it's complex technically, but nothing that almost one petadollar yearly budget can't solve [1]. [1] "White House Asks for $842 Billion in 2024 DOD Funding, 3.2 Percent Above 2023 Budget", https://www.nationaldefensemagazine.org/articles/2023/3/9/white-house-asks-for-842-billion-in-2024-dod-funding-32-percent-above-2023-budget https://www.nationaldefensemagazine.org/articles/2023/3/9/wh...
- mschuster91 3y agoThe thing is, do the US have a need for mine grids? You only need mines in a land war scenario as the Ukraine invasion shows at the moment, with Russian mines being the most pressing issue in the expected Ukrainian offensive campaign. The US in contrast, who should attack them with tanks or infantry? Canada and Mexico... both completely unrealistic, at least if one assumes that refugees are not supposed to be shredded by anti-personnel mines. And that completely leaves out the core problem with mines... it is easy enough to clear a path through a minefield using anti-mine weaponry, and Ukraine has gotten a lot of appropriate vehicles, ammunition and training. But the rest of the mines? They take many decades to clear up. When you go into the Croatian hinterland, where they used to fight off Serbians back in the 90s... you'll still see an awful lot of "warning, mine field" signs next to the road.
- nradov 3y agoThe US helps South Korea maintain a minefield in the DMZ. https://www.npr.org/2022/06/21/1106367928/us-landmine-ban-trump-korea-ukraine-russia-ottowa-treaty https://www.npr.org/2022/06/21/1106367928/us-landmine-ban-tr...
- nradov 3y agoIn theory sure. In practice US ground troops are already suffering frequent joint injuries due to the excessive weight of all the electronic equipment and spare batteries they have to carry. An AR headset robust enough for field use won't be light.
- HillRat 3y agoBlue force tracking has been a thing for decades now, so there's nothing new about (relatively) real-time unit tracking and C2 messaging in the field, and there have been multiple stabs at integrating HoloLens into the loop, none successful. (The last effort, Army's IVAS, got shut down by Congress due to reliability and usability problems earlier this year, though DOD hasn't given up on a less-immersive approach.) Where they've seen reasonable success has been in mission planning and simulation, and to a certain degree in technical training; the problems with moving XR systems from controlled spaces into the field have proven essentially intractable to date.