MISTT raises £5M to build a 4-D map of the world for AI and autonomous systems - Seedcamp

The Ordnance Survey was founded in 1791 because Britain’s maps could not keep pace with modern warfare.
Two centuries later, GPS did the same thing for the satellite age. Each technological shift has required a new reference system to help us make sense of the physical world.
MISTT are building the next-generation navigation system for AI and autonomous systems. We’re proud to invest in their £5M pre-seed round, alongside Expeditions, who led the round, Watchkeeper, the National Security Strategic Investment Fund (NSSIF), Overmatch and FJ Labs.
Defence is facing the mapping problem again, this time with data. Satellites, weather and ocean models, drones, radar and sonar each record position and time in a different format. Even the reference frames disagree, as Europe’s ETRS89 and the WGS84 frame used by GPS differ and drift further apart each year. As autonomous systems add more types of sensors, reconciling them is slow, manual work. Most existing systems flatten data into two dimensions, which impacts accuracy, especially in dynamic environments like the air or sea.
“Every new sensor adds to the complexity of the overall picture, as each understands the world from a different frame of reference,” said Dan Eaton, CEO and co-founder of MISTT. “Whether you are training a world model or planning across a targeting web, multi-domain data fusion is only possible if you have a strong data foundation.”
MISTT divides the world into a single grid of three spatial dimensions and one of time, from beneath the seabed to orbit. Every observation is given a single address in space and time, with uncertainty preserved as an asset. A back-end references and resolves very large datasets, so the data picture updates as datasets are ingested, and accuracy improves as more sources are added. MISTT is simplifying the data problem by taking on the biggest challenge of all: reconciling all data across all domains.
“At MISTT, we don’t build world models or autonomous systems,” Eaton said. “We are building the missing layer that provides them with accurate context of the world across space and time.”
The team has experienced this problem first-hand. Dan Eaton served as an officer in the Royal Marines for over 20 years before founding MIST. Chief Scientist Matt Bristow, formerly of DSTL, brings machine learning and geospatial data science expertise. CTO Casey Esworthy, chief architect at Microsoft until recently, brings a deep understanding of hyperscale architecture.
MISTT is starting with defence, the most demanding case, with a UK Ministry of Defence pilot beginning in December. It plans to expand across NATO and into the US, and is using the round to build its technical and business development teams.
Over time, the same data layer can serve autonomous vessels and drones, shipping, insurance and climate data, and AI world models that need anchoring in real-world data.
Seedcamp Investment Management LLP is a limited liability partnership incorporated in England and Wales with registered number OC391626 as authorised and regulated by the Financial Conduct Authority with FCA number 628313.