PolyChord, a London-based SME specialising in data science, has signed up to the national SPRINT business support programme to develop a unique Destination prediction and Estimated Time of Arrival (DESTA) system. SPRINT funding will enable PolyChord to collaborate with data science experts at the University of Southampton to progress the new DESTA technology that will better predict destination and arrival times of vessels carrying commercially valuable cargo.
Satellites are used to detect Automatic Identification System (AIS) signatures on ships when out of range of terrestrial networks, and information can be extracted to show unique identification, position, course, and speed. PolyChord will work with shipping and navigation experts from the University of Southampton to obtain more accurate information out of these data sets, and gain superior ETA and destination prediction. This can have a significant impact on the costs of goods as they move across the water and throughout the globe.
PolyChord’s innovative approach to the challenge of estimating the time of arrival is focused on using a more advanced kind of probabilistic inference than was previously available. The PolyChord Nested Sampling technology provides more accurate error bands for destination prediction.
The project will be funded by a grant from the £4.8 million SPRINT (SPace Research and Innovation Network for Technology) programme that provides unprecedented access to university space expertise and facilities. SPRINT helps businesses through the commercial exploitation of space data and technologies.
Mike Handley, Chief Executive Officer of PolyChord said: “The origins of the PolyChord technology are based on gaining information from complex and challenging satellite-derived data sets. We’ve developed a cutting-edge piece of data science technology that can help us to model areas more effectively by removing the guesswork and ‘quantifying the uncertainty’.
“However, we only have limited knowledge of shipping and navigation, and this is where the expertise of the University of Southampton is essential in this partnership. They bring expertise in machine learning and data learning, plus shipping and navigation data knowledge to the project. The interrogation and analysis of satellite data for commercial outcomes is where Southampton’s experience will be really valuable for business exploration.”
Adam Sobey, Associate Professor within Engineering and Physical Sciences at the University of Southampton and Co-Group Lead in Marine and Maritime within the Data-centric Engineering Program at the Alan Turing Institute added: “This project is all about taking satellite AIS data to predict behaviours of vessels; leveraging data science to identify arrival times for merchant shipping. Although this is a new area for the university, our background of working with ship data provides us with the necessary domain knowledge. We’re excited to test the capabilities of PolyChord’s dedicated tool and bring new methods to this domain from other disciplines.”
Picture caption: DESTA – using PolyChord’s unique data science technology to better track the destinations of ships with more confident predictions of ETA. Photo courtesy of NASA shows ships’ tracks revealed by satellite.