Skip to content

Latest commit

 

History

History
33 lines (20 loc) · 940 Bytes

File metadata and controls

33 lines (20 loc) · 940 Bytes

Indoor-Positioning-and-Navigation-using-Reinforcement-Learning

Basics

This project is used to solve the problem of indoor positioning. As gps does not work indoor, This project was tested for accuracy upto 15cm for indoor 2d positioning. The positioning this passed to a AI Model which navigates the vehicle to required position using Reinforcement Learning

Hardware

Three anchors are used in which UWB is connected to arduino pro mini The tag is connected to software serial of nodemcu i.e (D8,D7) Vehicle is controlled by Arduino Uno connected to NodeMCU Tx,Rx

The vehicle has gyro sensor and adafruit motor shield

Libraries

Add dwm1000 Custom Library : (Added custom channels and changed Tx Power) https://github.com/wahajmurtaza/arduino-dw1000

Add ESP Library: https://github.com/wahajmurtaza/NodeMCU-Python-Wifi

Python: pygame Pytorch Pyserial

Block Diagram