We're glad to see EE Times delving into the future of augmeted reality HUD, especially from a processing perspective, which is where it appears the sources for this article are based.
I want to respectfully suggest that not all solutions leading to a very wide FOV and appealing HUD lie with the chipmakers, however, although they play a central role in smoothing quickly rendered display of imagery used for navigation and ADAS HMI.
The company leading small form factor A/R HUD is MVS-California, in San Jose, California. Customers using that 3D, volumetric A/R HUD are beginning to expose their super wide FOV HMI plans to the world, through conferences and technical meetups. I notice none of those sources was quoted here, unfortunately, which is a shame because some of them are happy to let their researchers speak to the press.
Please go back to Strategy Analytics for additional information. They have recently updated a study on automotive display in which the state of augmented reality HUD is pretty thoroughly analyzed and its leaders identified.
It simply isn't possible to build a small in-dash A/R HUD by focusing exclusively, or even primarily, on better chips or SOCs, although those play a signficant supporting role. The optomechanic design leads, first and foremost. Projection and optomechanics are exquisitely archane fields in which chipmakers do not specialize.
NVIDIA and TI have every reason to be bullish on the future of large-scale, crisply defined Head Up Display. But the designs that will actually fit into a car, and those whose components will make such a design feasible, and affordable for average drivers, will come from specialist HUD design teams, not from chip makers whose business is spread across many competing display types. These entities have to work together, with chip makers following the lead of the HUD design teams.
Leaving distraction to driver aside, there are still tremedous challenges to projecting information on the windshield. The most obvious one will be how the information on the windshield being shown under bright sunlight. To overpower the brightness from the Sun is definitely not easy.
HUD? Hey, I still don't have keyless entry or an MP3 connection in my Honda CRV. My next car will hav that. I'm not sure I'm going to live long enough to own a car with an HUD. When does that come to a mainstream model?
I would expect the windshield itself to become the screen, eventually? No? Perhaps edge-lit. The windshield would be the matrix digital display.
Difficulties I can see include first and foremost, to decide what info to put on the windshield. Typically, it would be a very small amount of info. Anything more becomes a distraction. I can see your arrow idea, associated with GPS directions, makes sense. But to put a bunch of gauges up there doesn't make sense. It might be done as a sales gimmick, but it will end up being distracting.
The other problem is nighttime vs daytime brightness. At night, especially when driving on unlit roads, that windshield display has to be exceedingly dim. Otherwise, it will take away your night vision. Perhaps it would also need to be red, for that same reason. Imagine missing dark shapes in front of you, because some HUD object is blinding you to them.
Focusing distance may also be a problem, although maybe that's also a problem with regular instrument panels. You don't want people to spend too much time focusing on the HUD image in front of them, and simultaneously fuzzing out traffic around them.
Bert22306 makes excellent points about uncluttering the HUD view, with special sensitivity to night vision, maintaining necessary "driver-in-the-loop" acuity. We share many of these views. OEMs vary enormously in their treatment of augmented reality HUD, of course. Some quite conservative, others less so.
We have found that the 11 o'clock high position for TBT navigation is best, and most consistent with research done previously by NASA to help pilots navigate runways safely (and at night, maintaining peripheral vision without developing tunneling).
While I personally am not a fan of the "rip all content from the center stack and throw it into the HUD" approach, I imagine we will see a few of those in the early R&D stages. Once A/R HUD nears production (2016-17, trucks then cars), we may expect a consensus to have been arrived at around safe protocols.
One important quality of the very adept deep-field Augmented Reality HUD as developed by the MVS team is that it can portray images far, far in the distance, or nearer, to within 1 meter of location accuracy - and that window is getting tighter every year. So the aim is not to "create images" that don't belong in the driver's view, but preferbly - delicately, exquisitely - to simply light up, very subtly, those objects upcoming that a driver really *needs* to see.
For TBT navigation, we consistently recommend an 11 o'clock position, although the HUD can do "arrow on the ground" just as effectively. It's wonderful to hear such intelligent feedback on the subject.
This certainly will increase the cost of a windshield. I wonder by how much. I had to replace my windshield twice in the last 7 years due to stone damages. What if a stone hits my windshield: is the display damaged as well? How about ice in cold regions or direct sun exposure (UV) in hot regions? This may be a novelty for a long time until it becomes more lucrative for the average consumer.
Hi, Rick. Well, if you want HUD, you can actually buy those after-market HUD from Garmin or Pioneer. The one from Pioneer (based on DLP technology) looks very nice, but it still costs close to $1,000. Not cheap, for sure.
However, from what I understand, these HUDs are going downstream pretty quickly...you may get your HUD-preinstalled car by 2020.
@Bert, good points all around. Not to clutter HUD with too much information is definitely the key.
Some car makers are already saying that drivers can select what info to put up on the HUD.
The last thing I want to see carmakers putting in is a text msg from a driver's mobile on HUD.
But on the second thought, when peopel are getting in traffic accidents because they can't restrain themselves from taking a peak at text msgs on their phones, maybe, putting it up on HUD could save people's lives...
What are the engineering and design challenges in creating successful IoT devices? These devices are usually small, resource-constrained electronics designed to sense, collect, send, and/or interpret data. Some of the devices need to be smart enough to act upon data in real time, 24/7. Specifically the guests will discuss sensors, security, and lessons from IoT deployments.