An Ultra-Low Power Real Time Embedded System for Map Generation Using Ultrasound Sensors

In this paper, a low power structure for acquiring the range reading from a set of ultrasound sensors and calculating the probability of occupancy of cells in the region under the sonar scan is proposed. The architecture considerably lowers the switching activity at various stages of acquiring and processing the sensor data to provide updates of cell occupancy values for rapid in-motion mapping for robot navigation. Power reduction at various architectural levels have been attempted and simulation results show that the proposed architecture lowers the switching activity by over 70% and reduces the power consumption up to 60% as compared to the conventional fixed-point processor based architecture.