This paper describes a system architecture starting from the sensor design to the algorithm implementation to control a mobile robot using electrooculogram based eye tracking system. Various methods exist to control external devices with muscle movements, Electrocardiogram (ECG) or Electroencephalogram (EEG). The proposed system works with Electrooculogram (EOG) by sensing eye movement and is found beneficial for paraplegic/quadriplegic patients who can get assistance from a 2-DOF robotic arm to perform their daily tasks. This low cost design can be easily implemented on an embedded robotic platform for implementation in various domains.