Merge pull request #29 from wroscoe/dev

Add differential drive and update docs.
autorope · Feb 9, 2017 · e48ce77 · e48ce77
2 parents 1efcf96 + 45b4b81
commit e48ce77
Show file tree

Hide file tree

Showing 42 changed files with 1,200 additions and 634 deletions.
diff --git a/README.md b/README.md
@@ -24,7 +24,7 @@ Once you have built your car you can use it like this.
 
 3. Go to `<your_pilot_server_ip>:8887` on your phone or computer to start driving your car. 
 
- Use the [demos](demos) to see how to record driving data, train autopilots and more.
+ Use the [scripts](scripts) for more tools to manipulate datasets and train autopilots.
 
 
 

diff --git a/demos/drive_sim.py b/demos/drive_sim.py
diff --git a/demos/serve_no_pilot.py b/demos/serve_no_pilot.py
diff --git a/demos/serve_pilot.py b/demos/serve_pilot.py
diff --git a/demos/train_download.py b/demos/train_download.py
diff --git a/demos/train_generator.py b/demos/train_generator.py
diff --git a/docs/01-build_a_car.md b/docs/01-build_a_car.md
@@ -1,101 +1,3 @@
-# Get Started
-This doc will walk you through how to setup your donkey.
-
-
-## Bill of Materials.
-#### Required
-* Raspberry Pi 3 B ($35)
-* Raspberry Pi Camera ($15)
-* Micro USB power adapter. ($9)
-* USB Battery Pack ($15)
-* USB Keyboard 
-* USB Mouse
-* Monitor 
-* HDMI cable (to connect monitor to PI) ($7)
-* Micro SD Card (campatible with Raspberry Pi) ($8)
-* Servo Sheild ($15)
-* RC CAR (($120-300))
-
-#### Optional
-* Xbox 360 Controller
-* Xbox USB Adapter
-
-
-
-## Setup
-These instructions are based on [Geoff Boeing's post on setting up scientific python on a Raspberry Pi](http://geoffboeing.com/2016/03/scientific-python-raspberry-pi/).
-
-#### Boot your Raspberry Pi
-These instructions assume you're using Ubuntu operationg system. If you don't have Ubuntu, try using the NOOB method.
-
-1. Download recent Disk image. I use Jessie.
-2. Extract disk image from zip file. 
-3. Open Ubuntu's  "Startup Disk Creator" application. 
-4. Insert micro usb disk via a usb adapter. This disk should show up in the Startup Disk Creator app. 
-5. Select your RPi .img file and click create disk. 
-6. Once the img has been created, take the SD card from your computer and put it in your RPi.
-7.  Connect your Monitor with your HDMI cable, your mouse and keyboard and then finally power up the RPi by plugging in the Micro USB power adaptor. 
-
-
-#### Install Basic Libraries
-Since the RPi is not as powerful as a laptop, it can take a long time to install python packages (ie. numpy & PIL) using pip. Luckly Adafruit has precompiled these libraries into packages that can be installed via `apt-get`.
-
-1. Open a terminal (Ctrl-Alt-t) and upgrade your system packages.
-
-	```
-	sudo apt-get update
-	sudo apt-get upgrade
-	```
-2. Save These initial conditions.
-
-	```
-	dpkg -l > ~/Desktop/packages.list
-	pip freeze > ~/Desktop/pip-freeze-initial.list
-	```
-
-3. Install the necessary libraries 
-
-	```
-	sudo apt-get install xsel xclip libxml2-dev libxslt-dev libzmq-dev libspatialindex-dev virtualenv
-	```
-4. Pandas & Jupyter Requirements
-	```
-	sudo apt-get install python3-lxml python3-h5py python3-numexpr python3-dateutil python3-tz python3-bs4 python3-xlrd python3-tables python3-sqlalchemy python3-xlsxwriter python3-httplib2 python3-zmq 
-	```
-5. Scientific Python
-	```
-	sudo apt-get install python3-numpy python3-matplotlib python3-scipy python3-pandas 
-	```
-
-#### Install your Camera
-Follow the instructions [here](https://www.raspberrypi.org/learning/getting-started-with-picamera/worksheet/).
-
-
-1. Connect your camera to the RPi.
-2. Enable your camera in the Menu > Preferences > Raspberry Pi Config 
-3. Restart your Pi. `sudo reboot`
-
-
-#### Connect your servo sheild. 
-
-1. Assemble and test your servo shield with the instructions given by Adafruit. 
-
-
-#### Install Donkey
-
-
-clone repo & create virtual env
-
-```
-mkdir code
-cd code
-git clone http://github.com/wroscoe/donkey.git
-
-mkdir car
-cd car
-virtualenv --system-site-packages -p python3 env 
-source env/bin/activate
-pip install -e ../donkey[pi]
-```
-
+# Build a donkey vehicle. 
 
+Anyone can build their own robocar that runs Donkey. Once you have the parts it will take ~4hrs to start driving. See the [detailed instructions](https://docs.google.com/document/d/11IPqZcDcLTd2mtYaR5ONpDxFgL9Y1nMNTDvEarST8Wk/edit).
diff --git a/docs/02-start_driving.md b/docs/02-start_driving.md
@@ -16,4 +16,16 @@ Now that you have built your car, you'll want to drive it.
     sudo nmap -sP 192.168.1.0/24 | awk '/^Nmap/{ip=$NF}/B8:27:EB/{print ip}'
     ```
 3. Connect to your pi by running `ssh pi@<your_pi_ip_address>`
-4. Run the `drive_pi.py` demo script
+4. Activate your python virtual environment 
+	```
+	cd donkey
+	source env/bin/activate
+	```
+5. Start your drive script.
+	```
+	python demos/drive_pi.py
+	``` `
+
+
+#### Control your car
+You can now control your car with the virtual joystic on your computer or your phone.
diff --git a/docs/99-Contribute.md b/docs/99-Contribute.md
@@ -7,6 +7,18 @@ This is an opensource project to help accelerate the developement of self drivin
 3. Fork the code, make your change and submit a pull request. 
 
 
+###Guiding Developement Principles
+* **Modularity**: A self driving system is composed of standalone, independently configurable modules that can be connected modules.
+
+* **Minimalism**: Each component should be kept short (<100 lines of code). Each peice of code should be transparent apon first reading. No black magic, it slows the speed of innovation. 
+
+* **Extensiblity**: New components should be simple to create by following a template. 
+
+* **Python**: Keep it simple. 
+
+***These guidelines are nearly copied from Keras because they are so good*** 
+
+
 ## Tests
 Make sure all tests pass before submitting your pull request.
 

diff --git a/docs/datasets.md b/docs/datasets.md
@@ -0,0 +1,24 @@
+# Datasets
+
+Donkey has several builtin datasets to help users test their autopilots and confirm that they can learn from image sequences. 
+
+## Moving Square
+A single color moving square bounces around the screen.
+
+Outputs:
+* X - 120x160px image of moving square on black background.
+* Y - x and y cordinates of center of square in the image. (use options to only return x or y)
+
+
+
+## Driving Datasets
+
+### Line Following
+
+
+### Lane Keeping
+
+
+### Sidewalk
+
+### 
diff --git a/docs/Mechanical_Hardware_Instructions.md → docs/hardware/README.md b/docs/Mechanical_Hardware_Instructions.md → docs/hardware/README.md
diff --git a/hardware/Adjustable Camera Holder.stl → docs/hardware/camera_holder.stl b/hardware/Adjustable Camera Holder.stl → docs/hardware/camera_holder.stl
diff --git a/Hardware/Car Top Plate - Helion Conquest.dwg → ...hardware/hellion_conquest-base_plates.dwg b/Hardware/Car Top Plate - Helion Conquest.dwg → ...hardware/hellion_conquest-base_plates.dwg
diff --git a/hardware/Roll Cage 2nd Version.stl → docs/hardware/hellion_conquest-roll_bars.stl b/hardware/Roll Cage 2nd Version.stl → docs/hardware/hellion_conquest-roll_bars.stl
diff --git a/docs/hardware/hellion_conquest-top_plate.dwg b/docs/hardware/hellion_conquest-top_plate.dwg
diff --git a/Hardware/Car Top Plate - Trooper.dxf → docs/hardware/trooper-top_plate.dxf b/Hardware/Car Top Plate - Trooper.dxf → docs/hardware/trooper-top_plate.dxf
diff --git a/donkey/__init__.py b/donkey/__init__.py
@@ -1,17 +1,18 @@
 
 import os
 uname = os.uname()
-if uname[1] != 'raspberrypi':
+if not uname[4].startswith("arm"):
     from . import (utils, 
                    models, 
                    datasets, 
                    remotes, 
                    sensors,
                    actuators,
+                   mixers,
                    vehicles,
                    pilots,
                    templates,) 
 
 else:
     print('Detected running on rasberrypi. Only importing select modules.')
-    from . import actuators, remotes, sensors, vehicles
+    from . import actuators, mixers, remotes, sensors, vehicles
Original file line number	Diff line number	Diff line change
Expand Up		@@ -24,7 +24,7 @@ Once you have built your car you can use it like this.

		3. Go to `<your_pilot_server_ip>:8887` on your phone or computer to start driving your car.

		Use the [demos](demos) to see how to record driving data, train autopilots and more.
		Use the [scripts](scripts) for more tools to manipulate datasets and train autopilots.



Expand Down