You as developer must decide what the cleaner must do: search and identify trash to clean or just scan all the territory (rooms) to be sequentally cleaned. Both approaches involve image recognition algorithms and territory mapping algorithms.

 

You as developer must decide what the cleaner must do: search and identify trash to clean or just scan all the territory (rooms) to be sequentally cleaned. Both approaches involve image recognition algorithms and territory mapping algorithms.

I would prefer the second approach. Scan the whole room for garbage. What is the pattern recognition for this process? . Which algorithm of machine learning can be used? . Can CNN be used?

 

A robotic vacuum cleaner does not need RTOS.
This is a straight forward linear sequence.
Collision avoidance and detection is linear programming with no time constraints.

 

A robotic vacuum cleaner does not need RTOS.
This is a straight forward linear sequence.
Collision avoidance and detection is linear programming with no time constraints.

I agree with you but it need an intelligent software that can make the right decision. So the question is, is this an example of machine learning. I think yes, this is an example of machine learning.

Which algorithm of machine learning can be used? . Can CNN be used?

 

You can use a neural network for room map learning or make some simpler and less memory consuming algorithm yourself. Anyway you cannot sure your robot will be making the right decision. It is all about approximation and development cost (and usefulness)

 

You can use a neural network for room map learning

Is there any particular reason you chose neural-network?

I am interested to understand neural network for room map learning. In general I don't understand how it is like to train a model. I know that the model is a mathematical algorithm.

suppose the equation y = a + b were the values of the inputs a and b. If we know the value then we get the result y
3 + 5 = 8
How is it like to train a model as the input value is fixed and the output will depend on the input

What does it means to train robot for room cleaning

Anyway you cannot sure your robot will be making the right decision. It is all about approximation and development cost (and usefulness)

I have just taken an example to understand the machine learning concept, I am not going to make it now

 

I would not call this machine learning.
This is image recognition and mapping for the purpose of collision detection and avoidance, and perhaps path generation.

 

I would not call this machine learning.
This is image recognition and mapping for the purpose of collision detection and avoidance, and perhaps path generation.

@MrChips Self driving car is another example of machine learning.

I want to understand that once the program is running on the system, how does this program train a system?

Once the car is programmed, how will it train or learn on its own?

 

Two days ago I stood for 10 minutes and watched a robotic lawn mower do its thing on someone's lawn.
I could see that this robot uses the typical buried wire to determine the keep-out zone. Apart from that it uses a random stop-reverse-turn algorithm to back away from the keep-out zone. There is no AI or ML here.

I am in the process of designing and building my own robotic lawn mower. I plan on using a compass, accelerometers, laser beacons, and imaging in order to map out the lawn and create a systematic mowing pattern. There is still no RTOS, AI or ML employed here.

 

There is still no RTOS, AI or ML employed here.

@MrChips I agree that you use the most suitable algorithm for your project.

Have you developed RTOS, AI or ML based project for client? I am not asking for hobby project. If you have done AI or ML project we can discuss about it.

 

@Ian Rogers - Understood, and that's your choice. How fast did the encoder you mention need read? And what needed done with that result?

 

Hiya Boba.... the encoder was a bit of a one off... I had it on the slewing ring on a manitowoc...( Crane ) and the encoder was supposed to be a 16PPR which can get me a decent result.. This thing spins quite fast and I get a pulse every 600 us.. BUT!!! I couldn't get one and it was urgent.. The only encoder available was 64PPR... We were now at 150us... Easy peasy for a pic18..

However! I just placed a simple pic12f1840 on top of the encoder to do the pre-processing in asm then the Chip had time to process... Again!! However! I had already ditched the RTOS as the pic wasn't big enough for said RTOS.... Now I sport a pic32 so it would have probably given me more of a chance... All in the past now..

 

Here's an example of a RTOS...
In the course, for the final project, we used a webcam to take video of an analog clock with a Raspberry Pi. We then analyzed each frame to provide images of the second hand at each individual second, without any skips or blurred movement in the second hand, for 30 minutes. I found this to be pretty easy... I generally had 30 frames to work with. I did some image processing to determine when the hand of the clock didn't move much and saved that image. Another requirement was that you had to use at least two threads on the processor. You could do this all on one thread, but that defeats the purpose of learning about RTOS.
A poor persons Rolex:

For extra credit, we used a digital clock to take images every 100ms without any blurs. This was much more difficult - you needed to save 1 frame in 3 so you had to somehow synchronize your camera to your digital clock. I was unsuccessful in this endeavor, so I don't have a video to share, but I was close. It's a project I'll likely circle back around to when I have more time. Here's a few images to give you the idea.
You can clearly see the last digit from 600ms to 700ms is blurred, and would not be acceptable to complete the project. Only one student in our class was successful, and there were maybe 5 that were close. About 30 total people were in the class.

Frankly, this was the best academic course I've ever taken.

@tindel
I clearly don't understand what is the point of raspberry Pi OS in example. Pi OS is not a real time operating system. Video doesn't show comparison between Pi OS and real time operating system. Right now i have raspberry pi and pi camera so i can try this experiment. But still i don't have a clear idea what is happening in the video. so can you please tell me what exactly you have done.