UK: To compute a vessel’s position in littoral waters, the defence contractors Polaris Consulting have developed a software system that uses visual and radar data. What is interesting about this software system is that it is not dependent on GPS.
Money for this project was funded by the UK Ministry of Defence‘s Centre for Defence Enterprise. The system uses autonomous applications that allow self-navigating vehicles to get a fix on their location in coastal areas where an adversary is jamming GPS signals (“GPS denied environments”).
The common option for automated, GPS-free navigation is the inertial guidance system, long used in aerospace and defense applications. However, an inertial system’s estimate is based on dead reckoning from one initial position, and it degrades over time until the next fix.
It must be very sensitive in order to deliver an accurate estimated position, and the equipment is expensive; Honeywell’s contract for the U.S. military for installing and servicing its EGI system runs to $500 million.
Instead, Polaris’ software uses an algorithm for integrating input from optical and radar sources, analogous to a mariner’s triangulations in terrestrial navigation. It calculates a position based on what each input “sees” and then adjusts how much it trusts the data based on current conditions. If it’s foggy, the algorithm treats the camera-based position as less accurate.
The system has been tested in simulation only, but the results are promising: in computer-generated versions of Portsmouth Harbor and the Solent, the accuracy of its estimated position had an average error of about 100 feet from GPS-generated positions, and the deviation was almost always under 300 feet.
The Polaris Consulting team intends to make additional upgrades to improve accuracy – like making the system capable of using visual aids to navigation – and to test it on one of Autonomous Surface Vessels’ (ASV’s) self-driving small craft.