Software Enhanced GPS-free Air to Subsea
Abstract
Flare Bright is a UK small business and a developer of Inertial Navigation Systems (INS) in software, using standard MEMS sensors that are typically utilised by many drones, vessels and vehicles. Flare Bright has to date utilised its expertise in drones and we have used a Machine Learning Digital Twinning approach to model aircraft aerodynamics, control surface and propulsor performance in all areas of the flight envelope, which enables hyper-accurate Position and Navigation (PNT) for any drone type.Flare Bright has proven its technology and secured contracts from the US Department of Defense, the UK#s Ministry of Defence and some of the major defencePrimes to enable drones to fly in GNSS-denied environments.Flare Bright wishes to utilise this validated expertise in bringing thissoftware enhanced GPS free capability to surface vessels. This is most likely to be successfully achieved by first achieving subseaautonomy using Flare Bright#s techniques (we have changed the fluid we operate in from air to water). We then expect to fuse the learnings from the air and the sea to achieve the same in surface vessels, modelling the added complexity of waves, currents and winds.In this context, the overall project aim of this low Technical Readiness Level (TRL) science project is to determine how to achieveGPS-free autonomy using Flare Bright#s Machine Learning enhanced Digital Twin software techniques for surface vehicles. The projecthas been scoped out to span three years, with each year considered to be a follow-on phase of the project with the following overarching aims:# Phase 1 (Year 1) # Paper focused on techniques that are likely to work, and investigating what is the roadmap for our approach.# Phase 2 (Year 2) # Research how a subsea model would function, and what inputs and unknowns we would need to understand more for the modelling approach. This would include a proof-of concept subsea vehicle simulation and a key output of this would be an assessment of how the flow estimation accuracy is affected by the significant change in any underlying physics models used.# Phase 3(Year 3) # Surface model approach and utilising proof of concept simulation to get some initial results and quantify likely accuracy measures.We believe that each of these year long activities are achievable for approximately the same rough order of magnitude (ROM) cost of $65k per annum. More detail is provided for Phase 1 at this stage, and follow-on proposals will detail subsequent Phases when appropriate during Phase 1.
Document Details
- Document Type
- DoD Grant Award
- Publication Date
- Apr 11, 2024
- Source ID
- N629092412036
Entities
People
- Carl Sequeira
Organizations
- Office of Naval Research
- United States Navy