Uncrewed aerial systems operational risk analysis

Abstract

Uncrewed Aircraft present an ever growing number of use cases from remote sensing to logistics with projections for large numbers in the skies of the future. Segregating airspace especially for their operations is neither sustainable nor scalable particularly in and around urban areas with existing airspace and airport infrastructure. Integration of UAS into unsegregated airspace requires robust risk assessment in order to prevent exposure of existing airspace users to additional undue risk. This is referred to as "Air Risk".

A large number of use cases involve flight over urban areas, exposing third parties on the ground to additional risk from overflight. A majority of UAS do not require costly and stringent airworthiness certification akin to commercial passenger aircraft, therefore operational risk mitigation measures must be applied to ensure safety of operations to a suitable Target Level of Safety (TLS). This is referred to as "Ground Risk".

Quantitative and objective risk assessment methodologies specific to time and three dimensional space are developed for parametrised UAS, accounting for both spatiotemporal population movement in real world settings and aerial traffic patterns derived from surveillance data. Such methods allow for the evaluation of probability of undesirable events such as third party ground fatalities and mid air collision to form an overall value for the risk posed by the UAS at a given position and time. Evaluation for large areas with a given TLS enables the objective determination of safe regions for UAS operations.

As a probabilistic problem using primarily a sampling-based methodology with a large number of factors, a computational problem in terms of trade-off between resolution and computational time is present. Monte Carlo methods and Rare Event Simulation techniques are applied for estimation of air risk, whilst parallel computing and use of GPUs are shown to reduce computational time significantly for ground risk estimation.

The end goal is a holistic, objective and quantitative risk assessment methodology for determination of the safety of UAS operations.

More information

Identifiers

ORCID for Alex Pilko: orcid.org/0000-0003-0023-0300

ORCID for Andras Sobester: orcid.org/0000-0002-8997-4375

Catalogue record

Export record