EEVblog Electronics Community Forum
Electronics => Projects, Designs, and Technical Stuff => Topic started by: Georgy.Moshkin on April 24, 2022, 05:46:40 am
-
Hello,
I want to share some information about my project that might interest someone. This project is started around 2019. Attached PDF file is a brief analysis of existing solutions and my project introduction. More technical information and photos later in this topic. I see it as a baseball cap with planar module hidden in fabric. Main feature is real-time microwave sensing with 360° field of view, a 2D mapping through 8+ receive channels, and low cost. Processed information is transformed to visual/audio/haptic stimuli which is accessible for blind and/or deaf person.
-
This is a much much better application than what we did back in 2008 for battlefield use :-+
See:
https://patents.justia.com/patent/10551483
Our concept was to use the advanced chips being developed for MMW electronically scanned phased array radars into a Personal Area Radar integrated into the soldiers helmet with proper backside beam attenuation.
The information would be integrated into the night vision goggles for various modes of display like shown below and never occurred to us for use to assist the blind or partially blind, which is a great humanitarian use :)
Our goal was to improve the Situational Awareness for the solder, and we could encode the radar beams with information to allow Friend or Foe identification as shown in the Red and Green blotches superimposed on the Night Vision Image. Red meaning unidentified or potential Foe rather than Green for Friend. Much else was also involved but can't be disclosed.
Anyway, your application is a great use of these new small Phased Array Radar chips and technology!!
BTW it does work ;)
Edit: These are from a very old presentation which can't be shown in whole. Spelling.
Best,
-
phased array
Is that really necessary at mm-wave frequencies? (Say 24 GHz, because that stuff is cheap nowadays.)
Wouldn't it be possible to make (binary) zone plate lenses and image on individual patch antennas as pixels?
-
phased array
Is that really necessary at mm-wave frequencies? (Say 24 GHz, because that stuff is cheap nowadays.)
Wouldn't it be possible to make (binary) zone plate lenses and image on individual patch antennas as pixels?
I think it may be possible, but not quite sure what you meant by "zone plate lenses". Maybe approach you've described already exists (coded aperture radars, sparse MIMO 2D arrays, etc.).
Anyway, your application is a great use of these new small Phased Array Radar chips and technology!!
Actually, I use cheapest single channel 24GHz MMIC transceiver as local oscillator. Receiving mixers are based on low cost 12GHz FET transistors which perform well at 24GHz as passive resistive mixers. I face many dilemmas, important one is low level of available LO power for mixing. Another application
A small update: Early this year, I have applied to "HICOOL" competition. I hoped to win a money prize. Unfortunately, my project was not approved for entering the competition. My current plan is to do some more proof of concept prototyping and testing around January-February 2023.
-
Good luck with your project, seems very worthwhile.
On the horizon may be some new types of Phased Array Systems which your application might benefit from. Keep in tune with advanced technologies related to Integrated Phased Array Systems, the IEEE is a good source for such.
Best,
-
This might be a proverbial 'if all you know is a hammer everything is a nail' but this to me seems like a great application for D-band CMOS/SiGe radar chips. Doing such a thing as a one-man team is obviously not possible, but if you are interested in bringing this to the next level and commercializing this, I would say going that route might work. Either using flexible antennas with dies bonded directly to the flexible substrate, or perhaps a small laminate is a better approach (similar to how it is done for some 60GHz point-to-point ethernet links now).
Antennas at those frequencies would be on the order of a millimeter in size, so a small array might be able to fit on a 1x1 cm laminate.
You could even envision something like sensor fusion, combining the data from multiple radars or even having them all work coherently as one radar (though in that case I think you would need the entire thing to be ridgid, if your different sensors are constantly moving w.r.t. each other this won't work).
This is really cool!
I don't have a lot of time but if you ever need some feedback on some millimeter-wave aspects feel free to get in touch.
-
I think it may be possible, but not quite sure what you meant by "zone plate lenses".
Bit late, but I meant zone plate lenses (https://en.wikipedia.org/wiki/Zone_plate) ;) That said, astronomers also work with mmwave and it's actually possible to make mmwave lenses out of plastic, they use Ultra High Molecular Weight Polyethylene to make mmwave cameras, they just image the emissions from galaxies rather than reflected signals closer to home.
So instead of scanning a tight beam with a phased array and receiving with an omnidirectional receiver, you use an omnidirectional transmitter and receive with an array of rectennas behind a RF lens, detecting the timing of the received signal to determine depth. It's like a time of flight camera vs scanned LIDAR, but for RF instead of light.
Might need to use a modulated signal and some more signal processing behind the rectennas for sufficient SNR, simple threshold detection might not be enough. Complex, but a phased array with a sufficiently tight beam for high resolution is complex too.
It's a complex problem from which ever way you approach it, one of the reasons why LIDAR is still in a tug of war with time of flight cameras.
-
Maybe one thing to be aware of, at least in Europe: There is almost zero market demand for such solutions. Brace for quite some hostility as a for-profit venture going into this market sector.
Also, I don't see RADAR offering any advantages over cameras+object recognition and depth extraction.
-
Its like with cars, I suspect that the cost, durability, repair ability of such a device would be higher with a camera and software rather then MW hardware at the moment. But I am not sure. I think its gonna come down to someone on low ass government disability money having problems fixing the thing they need the most.
But none the less I would imagine that this solution might be superior because the algorithm might be simpler and you won't have problems with rain on optics etc? But you also might have problems with reflectors and interference that you won't have with optics.
-
If cars ever want to be really self driving in all weather, imaging mmWave will be necessary. A couple of radar zones isn't enough, you need near the resolution of LIDAR or a camera, but with something which can see through fog rain and snow.
Investors don't realize that, so you don't have to fake research into it to part them with their money though. Just keep talking about AI and employ AI researchers and the money keeps rolling in.
-
If cars ever want to be really self driving in all weather, imaging mmWave will be necessary. A couple of radar zones isn't enough, you need near the resolution of LIDAR or a camera, but with something which can see through fog rain and snow.
Investors don't realize that, so you don't have to fake research into it to part them with their money though. Just keep talking about AI and employ AI researchers and the money keeps rolling in.
Considering the amount of R&D that is being/has been poured into mmwave radar, and car manufacturers spooling up for 7+ mmwave radars in their next gen vehicle platforms, I think they are quite aware. 77 GHz is standard now, with 140 on the horizon.
-
I don't see a lot of research for high resolution 2+1D imaging for mmWave for the car industry. They seem to just use the higher frequency to get tighter beams, but still only a couple zones.
Only the security industry seems to do imaging.
PS. oops, seem to just have never noticed Arbe. Resolution is still a bit poor, but it's not zone based.