bluecobold / ankinodedrive

Anki Overdrive SDK using Node and BTLE

AnkiNodeDrive

Anki Overdrive SDK using Node and BTLE

This emulates much of what can be found in the default Anki Drive SDK:
http://developer.anki.com/drive-sdk/
This example shows similar capabilities using NodeJS.

Full documentation can be found in the apidoc/index.html; which is automatically served by the nodeJS server at '/'.

Running

There is a run.sh to show how to run the server using a specific BLTE device. I run this on a PI3 with built-in BLTE. If you only have one BLTE device, node server.js should be enough. You will likely need to do an npm install first to gather any dependencies.

Basic capabilities

See the apidoc for full api documentation. There are some APIs, such as set-sdk, connect, etc, which happen automatically, so you don't need to worry about them. For example, when you call "setSpeed", if the server is not connected to a car yet, it will automatically do a connect and turn on sdk mode. Also, logging is turned off by default, so you won't see messages from the car unless you explicitly call the 'turnOnLogging' service.

Here is a basic list of capabilities:
COMMUNICATIONS

• batteryLevel
• connect
• disconnect
• getDevices
• ping
• rescan
• turnOnLogging

DRIVE

• changeLanes
• setLaneOffset
• setSpeed
• trackCountTravel

LIGHTS

• flashTaillights
• setEngineLight
• turnOffHeadlights
• turnOffTaillights
• turnOnHeadlights
• turnOnTaillights

MAPPING

• getTrackMap
• getTrackMapData
• mapTrack

Example map generated

Known Limitations

I was only able to sort out the track IDs for my kit. I have a few other kits on order and will update the track IDs as best I can. Please let me know if you are able to tell me additional track IDs and what type of track it is (straight/curve/start).

The lights APIs were not easy to figure out. I believe they changed quite a bit in the Overdrive version and the public SDK does not address the changes yet. I did my best to handle a few cases. If you sort out others, let me know.

I began to work on being able to build a map of the track, but this is not yet complete. My thinking was to expose two services:
/mapTrack/:carname - This would drive the given car at a slow pace (300) and capture each track ID from start to finish building a track map array.
/getMap - Get the resulting map array which could be used to generate an image.
/getMapImage - Perhaps auto generate a png/gif of the track. TBD