vhoulbreque / ironcar Goto Github PK

Sometimes, captures are corrupted and cannot be opened.

XBOX controller :

• is not supported (no file/code to listen to the inputs)
• is still present in the README and the google docs...

There are 2 files :

• base_config.json (original configuration)
• config.json (the modified configuration, which is up-to-date, because modified by the user)

The goal would be to have only one file and a function (on the dashboard) to regenerate the original configuration.

Stays in the selected mode even when the user switches from main page to commands page.

Hello,

Unfortunately the installation does not work well apparently, since launching python3.5 ironcar_master.py gives the following error message:

File "/usr/local/lib/python3.5/dist-packages/scipy/linalg/blas.py", line 196, in
from scipy.linalg import _fblas
ImportError: numpy.core.multiarray failed to import

It looks like numpy modules are not found.

Could you help?
Many thanks!
Dominique

Add a button to test in 'live' the command values instead of having to switch back to /main constantly.

It can be a file or a section in the README

/home/pi/.local/lib/python3.5/site-packages/h5py/init.py:34: FutureWarning: Conversion of the second argument of issubdtype from float to np.floating is deprecated. In future, it will be treated as np.float64 == np.dtype(float).type.
from ._conv import register_converters as _register_converters

The users should be able to quickly retrain the model the day of the tournament with 'fresh data' from the circuit.

For example: direction
If right is too low or left too big, it can break the car...

To help people new to the project
To help people new to python/computer science/AI

I would like to know how do you find the commands defined in the json file.
The Adafruit controller can receive 16-bit messages (integers in [0, 4095]).
For exmples, these values can be sent to the steering motor : "left": 310, "right": 490.

But some messages are not interpreted by the gas or the steering motor.
For example, if I send 0 to the steering, nothing happens.
How can I find the complete dictionary {'value:action'} ? (where value is in [0, 4095])

My goal, find a finer set of commands for gas and steering!

Enable pictures to be taken even if stream is off

If straight, go fast
If turning, slow down

drive_max was used before to be sure that the car would be able to move forward. But now, it's easy to change the commands on the dashboard. So is it useful to stay with this value ?

It would be great to switch from Keras 1 to Keras 2, and update tensorflow versions accordingly.

Add automatic tests

Check if every necessary field is present (e.g. straight, dir_pin ...)

drive_max is faster in training than in autopilot (for the same value). Why ?

When the model has finished loading, there is no information telling it finished loading...

Python files are all in root, which is not good

See https://pages.github.com/

On Firefox, the "brain" emoji does not exist for instance

Or disable it

Add the wifi configuration in install.sh

https://medium.com/tensorflow/tensorflow-1-9-officially-supports-the-raspberry-pi-b91669b0aa0?linkId=55070050

On some cars, the drive value can change depending on the level of battery.
drive is high if the level is low
drive is low if the level is high

Sometimes, people don't have an available raspi (group projects with only 1 raspi for instance...)
So, they need to test the code on their laptop, but the only camera that is handled is the PiCamera...

• Constant gas
• Gas decreases when the car turns
• ...

Hello,
After running ./install.sh and launching node car_server.js with no problem, I have the following error message when I run python ironcar_master.py:

ImportError : No module named socketIO_client

It is true that I cannot find any py file with such name.
Did I do something wrong?
Thank you.

https://refreshless.com/nouislider/

When loading a model for autopilot, there is always this error (captured so it does not break the code):

The model ./models/autopilot_500k.hdf5 has been successfully loaded
Error on request:
Traceback (most recent call last):
File "/home/pi/.local/lib/python3.5/site-packages/werkzeug/serving.py", line 270, in run_wsgi
execute(self.server.app)
File "/home/pi/.local/lib/python3.5/site-packages/werkzeug/serving.py", line 258, in execute
application_iter = app(environ, start_response)
File "/home/pi/.local/lib/python3.5/site-packages/flask/app.py", line 1997, in call
return self.wsgi_app(environ, start_response)
File "/home/pi/.local/lib/python3.5/site-packages/flask_socketio/init.py", line 43, in call
start_response)
File "/home/pi/.local/lib/python3.5/site-packages/engineio/middleware.py", line 47, in call
return self.engineio_app.handle_request(environ, start_response)
File "/home/pi/.local/lib/python3.5/site-packages/socketio/server.py", line 360, in handle_request
return self.eio.handle_request(environ, start_response)
File "/home/pi/.local/lib/python3.5/site-packages/engineio/server.py", line 290, in handle_request
socket = self._get_socket(sid)
File "/home/pi/.local/lib/python3.5/site-packages/engineio/server.py", line 421, in _get_socket
raise KeyError('Session is disconnected')
KeyError: 'Session is disconnected'