Problem descriptions

Problems being lined up for the study group include:

  • A problem to develop climate crisis stress tests to assess corporate bonds.
  • Dstl – the science inside UK national security is considering the problem of uncertainty quantification of self-organised swarms of drone sensors.
  • Faraday Predictive Ltd want to understand the relative magnitude of distortions to motor current arising from changes in motor load, eccentricity and rotor condition. 
  • Innovation Embassy would like to estimate the total amount that online customers will spend based on limited data, including how the customer arrives at the website and their initial purchase.
  • Transfinite (problem 1) want to quantify risk under deployment uncertainty in radio spectrum management.
  • Transfinite (problem 2) would like to assess the probability of satellite collision and/or the number of avoidance manoeuvres that would be needed in busy orbit shells.
  • Vet AI want to develop a method to extend a medical coding system that minimises the risk of clinicians creating duplicate codes.
  • Zenotech would like to develop low dimensional/highly compressed approximations of point cloud data relating to aerodynamical hazards, such as regions of turbulence around buildings.