EOG signal interface based autonomous navigation of wheelchairs for disabled people

In this thesis, a smart wheelchair as an autonomous mobile robot for transportation and communication needs of disabled people who are unable to use their hands and feet has been developed. For this purpose; real time mapping, localisation, adaptive shortest path finding, autonomous navigation and control algorithms are developed and combined to transport the user to the desired destination point even if in previously unknown environments without any landmark and without commanding every step of the vehicle. Two prototype wheelchairs are developed: the first one works with Windows Operating System (OS) with encoder and Inertial Measurement Unit (IMU) based localisation algorithm. The second prototype works with Linux Ubuntu OS with encoder and LIght Detection And Ranging (LIDAR) based localisation algorithm without IMU. Destination points have been assigned with just one mouse click on the map of Graphical User Interface (GUI) by using electrooculography (EOG) and gyroscope signals of the user. Thus, the commands for transportation given by the user with EOG signals are reduced, became easier and practical. Besides this, as an additional feature, a special screen keyboard is developed to make computer easier to use and faster to write without using hands. New and smart wheelchair for disabled people is achieved with the developed system.