Hi everybody, i would like to show you my advance with the teleoperation topic, the main objective is use the Novint Falcon to operate the copter, avoid collisions and feel the velocities and distance to an obstacle by the force feedback. The firmware is for the pixhawk, based on some of the ArduCopter 3.2 libraries.
Also the copter is able to fly in external or internal environments, so if the gps signal is good to fly with it, it flies with gps, if not, i use the optical flow PX4flow to try to maintain the position.
To avoid the collision an ultrasonic sensor is used, but have some issues i think because the noise of the motors,
Still is not fully tested, but at this moment is functional. Greetings
Comments
Hi Sergio, I was wondering if you could join drones-discuss@googlegroups.com and have a peek at the "thoughts on obstacle avoidance integration with AC"? We've got a few people working on this but it's not clear how we should integrate this so your input would be greatly appreciated!
Hugues, in the first part with the strnage joystick (novint falcon) are send, the velocity commands that the copter follows, is like the loiter mode in Arducopter but with only one Joystick.
Hello , can you explain more what is shown on the first part with that strange joystick? What are the movement of the joystick translate?
Hi Mike, i´m also working with the maxbotix sensor using the PWM output, i didn´t find any other reason for the noise, more than the motors because when the motors are off the signal is solid lke you say, but when it´s flying it shows a measurement with noise, the firsts test the copter alway was moving backward and i didn´t know why until i check the log i realised of that, maybe i have a faulty sensor or the posistion affects the performance, multiple sensors sound great continue with your work, i would like to see that.
Laser, maybe the next step in this work is controlling multiple uav with one station.
I did some work on a prototype Quad with Sonar just over a year ago for collision avoidance in an inside environment. Interesting to see your comments on noise from the motors.
If you're using the an Analogue Voltage signal from the Sonar Sensor then that may well be an issue due to the electrical noise you could pick up. We used the Maxbotix sensors and at the time I stopped trying to filter out noise on the analogue signal and uses the PWM output and received rock sold distance readings. They also have an I2C bus connection.
I've since moved on to a production unit with multiple sonar using PWM at the moment, may use I2C in the future.
That is brilliant. What is the next step in developing this system?
Thank you Chris, i´m not working with the dev team, once Randy contacted with me for the implementation of the optical flow stabilization, i could implement the px4flow overriding one flight mode on Arducopter firmware and sent the link to Randy but that was all, before was not fully tested, but now after a lot of test i think is more stable.
Or implement what?
Love it! Great work. Hope you're participating with the dev team to integrate this...