Overview / Usage

Autonomous Driving is one of the most interesting AI problems, and Alpha Neuron aims to accelerate the world’s transition to driverless cars as the message of today’s car manufacturers is clear now; The cars of the future will be born fully autonomous.
Alpha Neuron consist of two parts the Autonomous Agent on top of a model car and Alpha Neuron interconnection protocol.
The model car is comprised of two sides hardware and software sides, since we are concerned with autonomy our main focus will be the software side of the project. Thus, for the hardware parts we will use prebuilt modules, The core modules for the project are: Deep learning modules which are responsible with the use of convolutional neural networks to detect lane lines, obstacles, street signs, traffic lights and to track other vehicles in the street, the second module is sensor fusion which make use of embedded sensors to get measurements and track pedestrians, the third module is the localization module which is responsible for locating the vehicle and planning a path in order to make the car navigate in the streets safely .

The interconnection protocol is a generic protocol for agents to communicate, the protocol will allow agents to organize traffic and increase safety further.

Hardware setup:
1/10 RC chassis
Raspberry PI 3
Rasspberry PiCam
Ultrasonic sensors
Movidius Neural Compute Stick
Potential Upgrades:
360-degree rplidar
Faster Replacement for Raspberry PI
More movidius NCS.
Intel Realsense cam.
2x current configuration for intercommunication protocol testing.