Good morning,
I hope this message finds you well. Lately I have been thinking about doing a project which will challenge me. Especially because I am very new in the field of electronics. My work is about a quadcopter drone that with LiDAR technology (mini TF sensor) can solve mazes and find exits inside. The idea would be to connect two mini TF sensors to an Arduino Original Rev3 board, which will be programmed and connected to the drone's flight controller. Being so new in this field, I wanted to ask you about your opinion of the project, and if the work is really feasible as I said it would be. Mostly I would like to know if the specified arduino, is enough to support my project, taking into account that the board will have possibly two sensors connected. I would also like to know how to connect the two sensors to an arduino UNO, or to a raspberry. And finally if anyone has an idea on how to bypass one of the two sensors, allowing the drone to solve mazes, and find exits in internal locations.
If you have any suggestions or comments, I would be happy to hear from you.
Thank you for your time and consideration.
Ricard Pozo
I imagine the mazes will be natural outdoor hedge mazes, in which case, your sensors will probably be a combination of visible and IR cameras.
If they’re man made paths, it’ll be easier with clear contrasting paths and boundaries visible from above.
Will not work. Take a look at ardupilot, it has LIDAR code built into it - but you'd need way more horsepower. If you have use for a RPiLIDAR https://www.slamtec.ai/product/slamtec-rplidar-a1/ drop me a note, I have one on the shelf that's collecting dust for a year now. With that thingie you could at least build a maze solving car.
That would work for a small maze, on a simple ground vehicle, and would be a good place to start your journey. Since this sort of thing has been a popular project for many years, plenty of Arduino example code is available on line for robot control, obstacle detection and maze solving.
As this is for a project, my only intention is that it simply works and I can at least get out of a room. You think an arduino UNO Rev3 is not capable of supporting it. If not, I also have the option of using a raspberry Pi, but that would mean spending more money, and I would like to know if with the arduino I have I would be able to achieve my goals. Then there would be the problem of how to program it, if anyone has any idea of how to help me I would be delighted.
Will the quadcopter be flying inside the maze or over the maze ? What material will the walls be made of and how wide will the passages be ?
My main idea would be for the drone to be able to find exits, possibly from a room or a building. I associate this with solving mazes, because at the end of the day, what the drone will be doing is going through spaces where there is only one way out. The drone will move around inside a room or corridors. Without controlling it, fully automated, only using both sensors, one at the front and one to the right.
Before tackling the maze solving you need to be sure that you can control the quadcopter's height and position in the maze corridor and can turn it accurately through 90 degrees
Have you thought how you are going to do that ?
The high would be constant, and since i don't want it to find a specific location, I just want it to find a exit fully automated. About the 90 degrees problem, yes, i've thought of a unique solution. The LiDAR will always be tracking the distance between itselve and the wall at a determined frequency of 1 sec, for example . In the moment, the LiDAR detects the distance is not constant, it will have to calculate a rotation and try to fix the error.
I appreciate that you want the height to be constant but how are you going to achieve that ? Keeping the motors at a constant power, even if you can achieve that, is not good enough
It isn't? what would you suggest to fix it. I didn't think about that
There are a couple of reasons why keeping the motors at a constant power is not good enough
Firstly, the quadcopter will be operating in a closed environment subject to air currents. This will be made worse by the maze walls causing turbulence. Secondly, the way that the quadcopter yaws is controlled by varying the power to diagonally opposite pairs of motors which in turn can affect the height
You can solve the second problem by having a mechanism to sense the height of the quadcopter and causing the motor power to be adjusted to compensate. To some extent this will also compensate for turbulence but only the vertical component. Turbulence may also cause violent horizontal movement too which will also need to be compensated for
You mention using 2 LIDAR sensors with a tracking frequency of once per second. This is way too slow. A quadcopter, even one moving at 5 mph will travel over 7 feet in a second. How wide is this maze ?
OK, do I need anything else, or are 3 sensors (bottom, front and right) enough? I was also wondering if the Arduino Rev3 would work for me, and if I would need a driver to connect all the sensors to the arduino.
In case of a separate suggestion, I would also be happy to receive.
How are you proposing to keep the quadcopter level ?
As you said, I would have to put a sensor in the lower part, to correct if the height varies.
Again: won't work. You don't have any expertise on building any airborn unit from scratch, not to mention a unit that's at least capable of maintaining it's position in 3D. These problems are hard. Go to ardupilot, enjoy some weeks of learning (and swearing and sweating).
You are confusing keeping the quadcopter at a fixed height with keeping it level. They are not the same thing
Imagine a quadcopter tilted at 45 degrees, and hence moving horizontally. It could still be maintaining its height
As has been suggested, you need to start by controlling the quadcopter's position in 3D. Only then can you consider moving it around a maze
Good morning,
I recently started a project for high school, which consists of building a quadcopter drone, and implanting LiDAR technology (distance sensors) so that it could go around rooms and find exits. As if the drone could solve mazes. I open this topic, in case someone could help me with the components for my drone. Since it is for a school project, I would prefer not to spend a lot of money. That's why, I'm addressing to this forum, with the aim of getting good components in terms of quality and price.
If you have any ideas that can help me, I would be delighted to hear from you.
Your two topics on the same or similar subject have been merged.
My main question is if the Arduino Rev3, is a good option for my project. To the arduino would be 3 LiDAR sensors Connected.