The Parrot: Sonar based multiple-direction ranging system

                            Home | Technology | Applications | Contact






During the operation, each Parrot node in turn acts as sender to drive the ultrasonic transmitters in the multiple ultrasonic sensor sticks to send out ultrasound simultaneously. The ultrasonic receivers in the neighbor Parrot nodes capture the ultrasonic signal, and the ultrasonic receiver that captures the first ultrasonic signal determines the direction of the sender.




Parrot nodes consist of a microcontroller, a radio transceiver and four pairs of ultrasonic sensors. Using ultrasound pulses to measure the distances between each other, each node in turn sends out a radio signal followed by an ultrasonic pulse train. Other Parrots that receive both signals convert the time difference between receiving the radio signal and the ultrasonic signal to a distance measurement by multiplying with the speed of sound.

The Parrots automatically formulate an ad hoc multi-hop wireless network to share the range information. They employ a distributed architecture in which every Parrot will have all the distance measurements for all the Parrots in the network, eliminating the requirement of a central node for the coordination. Every Parrot can be both sender and receiver. The Parrots coordinate the pining sequence such that ultrasonic pulses from multiple Parrots do not interfere with one another

Compared with similar ultrasonic ranging nodes such as the Cricket, the Parrot achieves a longer ranging distance of up to 15 meters with an accuracy of 2cm. It is also more effective in detecting ultrasonic signals by using four pairs of ultrasonic sensors, making it able to detect ultrasonic signal from any direction and reducing the number of nodes required by similar systems, which in turn leads to an improved update rate of the distance measurements and make the tracking of a moving object more efficient. Since the parrots can estimate the direction of the signal (to approximately 90 degrees) the ambiguity inherent to omni-directional range-only sensors is reduced, thus improving the accuracy of standard localization and mapping algorithms