... | ... | @@ -43,9 +43,10 @@ Try to play music and make mappings from the sound sensor to the two motors that |
|
|
As seen in the video the car moves forward when it detects a sound higher than 50 db. We have done this since the sound level that is detected by the microphone, when the room is totally quiet, differs between 10 and 20 db. We do not know exactly why this is, but could relate to a bad microphone or a too sensitive one.
|
|
|
We also notice that when the robot detects a decibel of eg. 60, the power that we have mapped to the motors are not enough for making robot drive. We did not manage to get a dynamic flow of the speed in relation to the distance of the sound source, it seemed like it only differed between three types om movement; backward, no movement, and full-power.
|
|
|
|
|
|
|
|
|
![Billed af afstand](http://gitlab.au.dk/uploads/u4099/legolabtimadala/3e8dd693b1/Billed_af_afstand.png)
|
|
|
##### Picture 1 Approximative distances - 0.5 cm. distance VS. 5 cm. distance
|
|
|
´´´
|
|
|
##### Picture 1 Approximative distances - 0.5 cm. distance VS. 5 cm. distance.
|
|
|
|
|
|
|
|
|
Another interesting aspect we found in regards to the detected db level by the microphone, was that when speaker was in almost as close as it could be to the sensor (see Picture 1), the db level was around 85, however when sound source was moved further away from the sensor, the decibel level increased to around 95. Again, we are not quite sure if this a result of weird behaving microphone or there are other factors or aspects that we have missed.
|
|
|
|
... | ... | @@ -100,4 +101,81 @@ The above code snippet shows the two different ways we have mapped the sound lev |
|
|
We managed to get the robot to behave to the environmental sounds. We have mapped these sounds to the power of the motor and experimented with using some of the values as negative values resulting in backward movement of the robot.
|
|
|
From the experiments we can see that the robot has a hard time reading values when the sound source is too close to the microphone. Furthermore is it hard to get a dynamic flow of the speed in relation to the distance of the sound source. This could be due to the quality of the microphone. Never the less gave this a inconsistency in the measurement when the sound source came too close to the microphone. The next time we look at this experiment it would beneficial to replace the microphone with another one, to see if the problem still exists.
|
|
|
|
|
|
## Exercise 2 - Vehicle 2; version 2a and 2b
|
|
|
### Goal
|
|
|
The goal with this exercise is to make an experiment, in which we make one of the vehicles Tom Dean mentions in [1] mounted with two light sensors in the front. We will use the vehicle 2a to experiment with both excitatory and inhibitory connections to see the different behaviors.
|
|
|
|
|
|
### Plan
|
|
|
For this exercise we will manly use vehicle 2a, but try to test the excitatory connection with the vehicle 2b. Our experiments will go as follows:
|
|
|
|
|
|
- We will connect the sensors to the motors like Tom Dean shows in with vehicle 2a see figure 1
|
|
|
|
|
|
![Braitenberg Vehicle 2a](http://gitlab.au.dk/uploads/u4099/legolabtimadala/259197bfad/Braitenberg_Vehicle_2a.png)
|
|
|
##### Figure 1 - Braitenberg Vehicle 2a
|
|
|
|
|
|
- We will first make the vehicle search for a dark garage through an excitatory connection. Here we will make the the vehicle turn away from the light and hide in the garage.
|
|
|
|
|
|
- We will after this make an experiment that uses the Vehicle 2b as seen in figure 2. Here we will test the behavior with an excitatory connection.
|
|
|
|
|
|
![Braitenberg Vehicle 2b](http://gitlab.au.dk/uploads/u4099/legolabtimadala/8ecc94c182/Braitenberg_Vehicle_2b.png)
|
|
|
##### Figure 2 - Braitenberg Vehicle 2b
|
|
|
|
|
|
- In the next experiment we will make the vehicle 2a with an inhibitory connection. Here the vehicle should drives towards the light and stops when there is too much light.
|
|
|
|
|
|
- Tom dean MAX_LIGHT and MIN_LIGHT experiment.
|
|
|
|
|
|
- Lastly we will explain and reflect upon video [9] - Vehículo de Braitenberg Sensores Ultrasónicos.
|
|
|
|
|
|
|
|
|
### Results
|
|
|
All the experiments were made under as similar light condition as possible as seen in picture 2.
|
|
|
|
|
|
![The light conditions.](http://gitlab.au.dk/uploads/u4099/legolabtimadala/bd5ccf9e0b/The_light_conditions..JPG)
|
|
|
##### Picture 2: The light conditions.
|
|
|
|
|
|
####Vehicle 2a- Excitatory
|
|
|
We have made an experiment where the vehicle finds a garage, as seen in Video 3 below.
|
|
|
|
|
|
[![image alt text](http://img.youtube.com/vi/4LKgasz8anE/0.jpg)](http://www.youtube.com/watch?v=4LKgasz8anE)
|
|
|
##### Video 3 - Video of vehicle 2a finding its garage - excitatory connection.
|
|
|
|
|
|
The video shows that we had to help the robot a bit finding its way into the garage. This is probably because of the lighting conditions in the room in which we performed the experiment, as seen in picture 2. We experimented with a direct light source trying to make the garage on the opposite side seem darker, but the vehicle still had some trouble finding the garage by itself.
|
|
|
|
|
|
|
|
|
####Vehicle 2b - Excitatory
|
|
|
In the experiment with vehicle 2b, we used the same code as in the experiment with vehicle 2a, but the results were a lot different.
|
|
|
|
|
|
[![image alt text](http://img.youtube.com/vi/PZUEhHuiW0I/0.jpg)](http://www.youtube.com/watch?v=PZUEhHuiW0I)
|
|
|
##### Video 5 - Video of vehicle 2b with an excitatory connection.
|
|
|
|
|
|
As seen in Video 5 the vehicle (2b) accelerates towards the light instead of turning against it. It is still only stopping when there is no light and are therefore continuing through the light.
|
|
|
|
|
|
|
|
|
|
|
|
#### Vehicle 2a - Inhibitory
|
|
|
[![image alt text](http://img.youtube.com/vi/8DpxHx9AwUU/0.jpg)](http://www.youtube.com/watch?v=8DpxHx9AwUU)
|
|
|
##### Video 4 - Video of vehicle 2a with an inhibitory connection
|
|
|
|
|
|
In the video we see that we achieved a better result in this experiment than in the previous one. The vehicle drove towards the light source and stopped right in the middle of it. We still had the same light conditions and settings as in picture 2.
|
|
|
|
|
|
#### Vehicle 2a - Max-Light and Min-Light experiment
|
|
|
|
|
|
We have in this experiment tried to show how the different rate of N samples influents the vehicles behavior. As seen in video 6 where N=10, which gives the vehicle a more fluent behavior.
|
|
|
|
|
|
[![image alt text](http://img.youtube.com/vi/otqaDsVZj0s/0.jpg)](http://www.youtube.com/watch?v=otqaDsVZj0s)
|
|
|
##### Video 6 - video of vehicle 2a with a rate of 10
|
|
|
|
|
|
In Video 7 we see a slow vehicle. Here N is set to 100. The reason for the slowly behaving vehicle could be due to the amount of samples it has to compare to the new value before performing a new activity.
|
|
|
|
|
|
[![image alt text](http://img.youtube.com/vi/uJr25tRH6FQ/0.jpg)](http://www.youtube.com/watch?v=uJr25tRH6FQ)
|
|
|
##### Video 7 - Video of vehicle 2a with a rate of 100
|
|
|
|
|
|
#### Video [9] Explanation/Reflection (two ultrasonic sensor)
|
|
|
In the video, Braitenberg’s vehicle 2 is used with two ultrasonic sensors, you see that the robot bumps into the second chair on its journey. This could relate to the angle that the sensors have towards the chair, here the ultrasound is not reflected back to the sensor and therefore the robot does not make the turn. We are not sure why the problem with the wall occurs, but perhaps it could be due to the reflection from the wall that reaches both sensors and some kind of interference happens. The sensors might be in continuous mode which results in the sensors not being able to distinguish the different beeps from each other. Here It would be more beneficial to use the ping mode, to avoid simultaneously echos from the sensor.
|
|
|
However, it could also be due to the angle towards the wall that makes it hard for the robot to find the right way. The solution to the problems could be to rebuild the robot, by placing the sensors closer to each other. You could also place the sensors further away from the floor, to avoid any interference from reflections. Lastly if the sensors was working in ping mode, as mentioned, it would prevent two pings from being activated at the same time and thereby avoid disturbance.
|
|
|
|
|
|
### Implemented code:
|
|
|
The implemented code can be found in reference 12.
|
|
|
|
|
|
### Conclusion
|
|
|
In the process of conducting the first experiment the robot had some problems when trying to find the dark garage. When conducting the next experiment, finding the light, it worked better. It drove towards the light and stopped right in the brightest light. In the video[9] that used two ultrasonic sensor to navigate the vehicle, which resulted it to bump into obstructions. It looked like it was using the continuous mode for sensors and poor angles resulting in the bad results. |
|
|
\ No newline at end of file |