Project Proposal by Martin Stacey


Route planning for autonomous flying taxis

Software

An object oriented language, possibly a genetic algorithm toolkit.

Covers

AI planning, autonomous vehicles

Skills Required

Interest in artificial intelligence

Challenge

Conceptual ??? Technical ??? Programming (????)

Brief Description

The rapid development of autonomous vehicles that can cope with traffic, and the development of drone technology creating aircraft and helicopters than can fly themselves has created the possibility that autonomous flying taxis might leave the pages of science fiction and become reality. Projects to develop autonomous passenger aircraft that could operate as taxis are already in progress, though it's less clear whether or when they will be able to carry useful payloads: See for instance the Munich-based company Lilium, which has carried out test flights of a VTOL personal aircraft that is designed to carry 200kg.

But how is the transport infrastucture going to work? Where are the airborne taxis going to go, and where shouldn't they go, and who or what is going to control this? How are collisions to be avoided and how will airborne traffic operate efficiently? How can unpredicted events be worked around? (For complex scheduling problems, such as airline flight scheduling, often the big difficulty with handling disruptions is the time involved in doing dynamic replanning.)

Your challenge in this project is to develop a concept demonstrator prototype route planning system that solves as many of the issues as possible, to handle a version of the air taxi management problem that is as realistic as possible, plus a demonstration environment that shows it working.

Variants

This project could have a number of different emphases, resulting in rather different projects.

Figuring out conceptually what is involved in managing airborne passenger traffic would be a rather different project: see Regulating and managing autonomous flying taxis


Back to