My name is Oyvind and I am one of the founders of Intelligent Agent (www.intelligentagent.no). We are a startup that develops radar technology for robotics.
We want to make radar technology available for UAV developers and are considering the possibility of offering a radar altimeter. We have most of the technology readily available and won't need too much time to finish a working prototype.
To get some input, I would like to start a discussion here to hear from you guys what your thoughts are on this.
What do YOU use an altimeter for? To keep a certain distance above ground? For autolanding? For avoiding collision with ground? Other?
And what is important for this application. Is it long range? Is it accuracy? Resolution?
I'd be interested in an altimeter for coping with varying terrain. What do you think is possible? It would need to work at altitudes of 100m or more, and would need to use only a minimal amount of power (maybe 2W would be ok?). It would also need to be light and cheap! Is that possible with current radar technology? What sort of radar frequency would it use?
If we were to put together what we have today, it would probably consume 3.5W, weigh about 200-300 grams, with a range up 60 meter.
Altitude of 100 m and 2W power consumption might be possible, but would require more development time.
Price is hard to say without knowing how many we could sell. But let us say that we knew we could only sell about 100 units a year, then maybe we could manage to push the price down to $1000 per unit. But if we knew we could sell 5000 a year, maybe we could get the price down to $500.
When you say you want to cope with varying terrain, does that mean that you for example would like your drone to always stay at a certain distance over the terrain? In which situations is this important?
yes, staying at a fixed height above terrain is the application I'm thinking of. There are a few reasons for this.
One is that there are regulations on the altitude above ground level that a UAV is allowed to fly. For example, a limit of 400 feet is common in Australia.
The second is to cope with rapidly changing terrain, so the UAV doesn't crash into it! Some people fly ArduPlane in mountainous regions. A downward pointing radar won't help with a cliff, but it would help with terrain that rises more slowly.
Finally it would be useful for aerial photography. Knowing the altitude is important for accurately georeferencing photos and recognizing ground features.
A price of $1000 would probably be out of the affordable range for most amateur UAV operators though. For my own project I'd probably just rely on the SRTM data that we use now.
Thanks for your insights!
What is the price-threshold that would make amateur UAV operators switch to radar altimeter?
Personally I believe it should be between $400 and $600 no more
What has put me off trying autolanding is accuracy. I fly gliders with large airbrake flaps. It is critical to get the flaps up from full brake position during final flare otherwise they tend to break the servo as the flap hits the ground. My estimated requirement is +-10cm or better accuracy over the closest meter. I might suggest a 30m range with +-10% accuracy to enable well controlled final approach.
If you wish to have the majority or people buying then it needs to be cheap enough to loose/damage and not have sleepless nights over it. Having said this, it is not a standard piece of consumer equipment like a gps. I would expect it to cost more. Should the radar be the next piece of equipment on the UAV enthusiast wishlist after they get their first UAV flying?
If you consider the latest 3DR gps at $75. https://store.diydrones.com/3DR_GPS_LEA_6_p/br-3drlea-6.htm
and an FPV setup at ~$250+ just for the goggles. Too expensive and people will go for the FPV as a more fun option.
Question: Would this radar module be mountable inside the aircraft or does the antenna need to go outside?
A couple of design considerations, if used for planes to hold a predefined altitude above terrain it should be tilted forward instead of vertically pointed to the ground. Also it should be able to operate if the plane is in a +-20 degrees roll position .
I've been working for a while on scaling down an opensource RADALT built by a cool Slovenian dude, I believe it will be a good start point for you too.
Interesting article. Have you calculated the expected component cost for this?
Crashes and fly-away losses are common enough for everyone to be thinking of it. We still go and put an extra $200 of autopilot onboard.
I suspect that putting an extra $400+ of radar onboard is not going to fit most budgets + risks. Having said that, DIYD is 15,000+ members so you have a fair market to work with.
I don't have a feel for how much $$ cost there is in the microwave components for the radar. The remainder of the RADALT blobk diagram looks similar in complexity to the APM. Therefore you might target the APM cost or below as a benchmark.