4 ms·

I was wondering the same thing myself today. It'd be interesting to come up with a drone that's ideal for this purpose… that can quickly fly and relocate somewh

I was wondering the same thing myself today. It'd be interesting to come up with a drone that's ideal for this purpose… that can quickly fly and relocate somewhere, but can land on any flat surface and record remotely. You could perch it on nearby buildings or even things like traffic lights

samizdis 6y ago
Defence companies have been touting "perch and stare" capabilities for quite a while. This is from 2017, but gives the gist of the idea: https://defensesystems.com/articles/2017/09/cw/army-research-lab-drone.aspx https://defensesystems.com/articles/2017/09/cw/army-research... Edited to add: this is the sort of thing: https://www.pasadenastarnews.com/2017/05/22/aerovironments-cell-phone-size-drone-is-helping-us-military-forces/ https://www.pasadenastarnews.com/2017/05/22/aerovironments-c...
tonyarkles 6y ago
As a warning for people who might try to do this with existing off-the-shelf drones, be warned that trying to do this is likely going to result in you losing your drone. Not that you'll have a crash, but my experience with a wide range of drones (from little $30 whoops to $40k commercial cinematography units) is that they often do not like to takeoff again after landing and sitting for any length of time. Best I can tell, accelerometer/gyro/compass errors start to accumulate while the aircraft isn't moving, and it puts the flight controllers in a bad state. I actually had a client ask me about just such an application yesterday. It's a really great concept and definitely worth exploring.
- gruez 6y ago
  That doesn't make any sense. How is the drone supposed to take off during normal operation?
  - leetcrew 6y ago
    GP is saying the sensors get into a bad state if the drone is powered on but not flying for a while. "normal operation" is powering on the drone and taking off almost immediately, then landing after 15-20 minutes. it's not inconceivable that most testing would follow this "happy path". the battery life of an idling drone might exceed the length of most test cases.
    - gonzo41 6y ago
      This is where a drone platform with an OS driver / control software would be great. I bet the sensors and the control systems all have memory overflows or number range issues in the software that isn't being dealt with because it's a closed system. Add to that few if any customers are asking to use their drone as a static tripod.
      - tonyarkles 6y ago
        I'm planning on doing some experiments with that after we're done spring surveying. To start: I'm not sure what you mean by OS driver/control software. The two main (non-proprietary) flight controller platforms are PX4 and ArduPilot, which both basically act OS-like. Depending on which platform you're building for, they can use things like ChibiOS under the hood as well. It's not memory overflows. These can fly for long periods of time, but sitting on the ground for a couple of minutes can be enough to require a reboot. What it is, best I can tell from the limited debugging I've done so far, is accumulated sensor bias. When the aircraft is in the air, you've got some sensors providing short-term data (accelerometers, gyros, magnetometers) and some sensors providing long-term data (GPS, optical stabilization if fitted); the short-term sensors are noisy, and an Extended Kalman Filter does sensor fusion to provide estimates that are the combined values from the fast sensors and the slow sensors. When a GPS update comes in, the EKF corrects for the biases in the sensors. When you're sitting on the ground in a stationary position, you're only measuring noise. The accelerometers are saying you're accelerating a little bit this way and a little bit that way; the gyros are saying you've spun a bit; the GPS says you've moved around some. The GPS signal is nominally mean-zero motion, but the rest of the sensors give both bias and noise. The EKF tries to filter out the noise and correct for the bias, but because the GPS is also wobbling around, it seems that it doesn't do the bias estimation correctly. For example, this sample, the accelerometers say that you've accelerated south, but the GPS sample has moved 0.5m north. Also, GPS doesn't help correct everything unless the aircraft is in motion; if there's magnetometer drift and gyro bias in the same, the aircraft has no way of knowing that it hasn't rotated (when you're airborne, you can use the GPS signal to know which direction you're moving, even if the magnetometer says you're pointed off by a few degrees).
aendruk 6y ago
A fun variation on the idea is to cling to rough vertical surfaces like an insect: https://spectrum.ieee.org/automaton/robotics/drones/stanfords-flying-perching-scamp-can-climb-up-walls https://spectrum.ieee.org/automaton/robotics/drones/stanford...

samizdis 6y ago

Defence companies have been touting "perch and stare" capabilities for quite a while. This is from 2017, but gives the gist of the idea: https://defensesystems.com/articles/2017/09/cw/army-research-lab-drone.aspx https://defensesystems.com/articles/2017/09/cw/army-research... Edited to add: this is the sort of thing: https://www.pasadenastarnews.com/2017/05/22/aerovironments-cell-phone-size-drone-is-helping-us-military-forces/ https://www.pasadenastarnews.com/2017/05/22/aerovironments-c...

tonyarkles 6y ago

As a warning for people who might try to do this with existing off-the-shelf drones, be warned that trying to do this is likely going to result in you losing your drone. Not that you'll have a crash, but my experience with a wide range of drones (from little $30 whoops to $40k commercial cinematography units) is that they often do not like to takeoff again after landing and sitting for any length of time. Best I can tell, accelerometer/gyro/compass errors start to accumulate while the aircraft isn't moving, and it puts the flight controllers in a bad state. I actually had a client ask me about just such an application yesterday. It's a really great concept and definitely worth exploring.

gruez 6y ago
That doesn't make any sense. How is the drone supposed to take off during normal operation?
- leetcrew 6y ago
  GP is saying the sensors get into a bad state if the drone is powered on but not flying for a while. "normal operation" is powering on the drone and taking off almost immediately, then landing after 15-20 minutes. it's not inconceivable that most testing would follow this "happy path". the battery life of an idling drone might exceed the length of most test cases.
  - gonzo41 6y ago
    This is where a drone platform with an OS driver / control software would be great. I bet the sensors and the control systems all have memory overflows or number range issues in the software that isn't being dealt with because it's a closed system. Add to that few if any customers are asking to use their drone as a static tripod.
    - tonyarkles 6y ago
      I'm planning on doing some experiments with that after we're done spring surveying. To start: I'm not sure what you mean by OS driver/control software. The two main (non-proprietary) flight controller platforms are PX4 and ArduPilot, which both basically act OS-like. Depending on which platform you're building for, they can use things like ChibiOS under the hood as well. It's not memory overflows. These can fly for long periods of time, but sitting on the ground for a couple of minutes can be enough to require a reboot. What it is, best I can tell from the limited debugging I've done so far, is accumulated sensor bias. When the aircraft is in the air, you've got some sensors providing short-term data (accelerometers, gyros, magnetometers) and some sensors providing long-term data (GPS, optical stabilization if fitted); the short-term sensors are noisy, and an Extended Kalman Filter does sensor fusion to provide estimates that are the combined values from the fast sensors and the slow sensors. When a GPS update comes in, the EKF corrects for the biases in the sensors. When you're sitting on the ground in a stationary position, you're only measuring noise. The accelerometers are saying you're accelerating a little bit this way and a little bit that way; the gyros are saying you've spun a bit; the GPS says you've moved around some. The GPS signal is nominally mean-zero motion, but the rest of the sensors give both bias and noise. The EKF tries to filter out the noise and correct for the bias, but because the GPS is also wobbling around, it seems that it doesn't do the bias estimation correctly. For example, this sample, the accelerometers say that you've accelerated south, but the GPS sample has moved 0.5m north. Also, GPS doesn't help correct everything unless the aircraft is in motion; if there's magnetometer drift and gyro bias in the same, the aircraft has no way of knowing that it hasn't rotated (when you're airborne, you can use the GPS signal to know which direction you're moving, even if the magnetometer says you're pointed off by a few degrees).

aendruk 6y ago

A fun variation on the idea is to cling to rough vertical surfaces like an insect: https://spectrum.ieee.org/automaton/robotics/drones/stanfords-flying-perching-scamp-can-climb-up-walls https://spectrum.ieee.org/automaton/robotics/drones/stanford...