Carputer is a 1/10th scale self driving car. It drives based on video camera and speedometer inputs. It uses a neural network to drive. The camera image and speedometer get sent to the neural network and the network outputs steering and throttle motor controls for the car.
Since the car uses a neural network for driving, it has to be trained. The training process is basically driving it around a track about 20 times or so using the car's radio control. During that driving, we record all of the radio control inputs along with the video and speedometer data. Once we have that, the neural network can learn how to mimic our driving style by outputting steering and throttle based on the video and speedometer.
The process is to record, train, and then run autonomously, as seen in the steps below.
This is an experimental setup, so it's not super-clean code or hardware.
- Turn on ESC, RC controller. Plug in battery, USB. Start switch to off.
- Run InsomniaX and disable lid sleep and idle sleep.
- activate the virtualenv:
source /path/to/venv/bin/activate
- run a script to drive and record training data:
python main_car.py record
-- this will let you have manual control over the car and save out recordings when you flip the switch
- Turn on ESC, RC controller. Plug in battery, USB. Start switch to off.
- Run InsomniaX and disable lid sleep and idle sleep.
- activate the virtualenv:
source /path/to/venv/bin/activate
- run a script to let tensorflow drive:
python main_car.py tf
-- when you flip the switch, you will lose manual control and the car will attempt to drive on its own - for autonomous kill switch: pull throttle and turn the steering wheel
- to revive autonomous mode, hit the channel 3 button (near the trigger)
- convert TRAINING images to np arrays:
python NeuralNet/filemash.py /path/to/data
(Can be multiple paths) - convert TEST images to np arrays:
python NeuralNet/filemash.py /path/to/data --gen_test
(Can be multiple paths) - train a model:
python NeuralNet/convnet02.py
. Train for minimum 1500 iterations, ideally around 5000 iterations. - use this model to drive the car (see above)
- for training info, see
debug.html
-- reminder: < 7 is right, > 7 is left - run
analysis/make_video.py
for debug videos - use
analysis/plot_vs_time.py
to view telemetry data
- Fix the radio control dropped signal error
- Get the TX1 working
- Get the IMU recording data
- Fix keepalive on Arduino
- Look into remote SSH type software so we don't have to keep popping open the car.
Updates - We no longer use the IMUs and we're no longer trying to run the NVidia TX1 computer. Macbook works better.
Work in progress, of course. The simulator runs in Unity. I didn't check in the lighting files because they are big, but if you build lighting, it should look like this...