I am interested in developing an intelligent autopilot system which not only can
perform full flights safely and autonomously from one location to another, but also
capable of handling flight uncertainties such as severe weather conditions
and emergency situations. The proposed Intelligent Autopilot System (IAS) learns
how to perform piloting tasks and the required skills from human teachers
using different aircraft, by applying the learning by imitation concept with
Artificial Neural Networks, and using an approach that requires few examples
and small training datasets, which speeds up development ,enhances performance,
eliminates the black-box issue, and makes comprehensive V&V possible.

The IAS can either be applied as an advanced digital co-pilot which a human captain
can completely rely on, or a fully autonomous pilot that can fly on its own. The IAS can
increase safety by tackling issues such as the low capabilities of modern autopilots,
stress, information overload, emergencies, and sometimes the lack of sufficient
and up to date training. The IAS can also lower costs associated with the need
for more than one pilot in the cockpit, or ground control facilities and staff
in the case of Unmanned Aerial Systems.


Coming soon.. H. Baomar and P. J. Bentley, "Training and Evaluating the Intelligent
Autopilot System for Full Flight Cycles with Oman Air, and Landing Autonomously Beyond
the Current Limits and Capabilities Using Artificial Neural Networks", 2019

H. Baomar and P. J. Bentley, "An Intelligent Autopilot System that learns Piloting Skills
from Human Pilots by Imitation", 2016 International Conference on Unmanned Aircraft
Systems (ICUAS), Arlington, VA, USA, 2016, pp. 1023-1031.

PDF   link

H. Baomar and P. J. Bentley, "An Intelligent Autopilot System that Learns Flight Emergency
Procedures by Imitating Human Pilots", 2016 IEEE Symposium Series on IEEE
Symposium Series on Computational Intelligence (SSCI), Athens, 2016, pp. 1-9.

PDF   link

H. Baomar and P. J. Bentley, "Autonomous Navigation and Landing of Airliners Using
Artificial Neural Networks and Learning by Imitation", 2017 IEEE Symposium Series on
Computational Intelligence (SSCI), Hawaii, USA, 2017.

PDF  link

H. Baomar and P. J. Bentley, "Autonomous Landing and Go-around of Airliners
Under Severe Weather Conditions Using Artificial Neural Networks", The 2017
International Workshop on Research, Education and Development on Unmanned
Aerial Systems (RED-UAS), Linköping, Sweden, 2017.

PDF  link



A complete flight in calm weather - cockpit view

A complete flight in extreme stormy weather - cockpit view

Extreme crosswind landing

A step towards the future..   Artificial Horizon LLC


15/09/16 - The Economist - Flight response


16/09/16 - The University Paper - UCL experts develop autopilot AI that learns on its own


29/03/17 - WIRED - AI Wields the Power to Make Flying Safer—and Maybe Even Pleasant


18/04/17 - BBC Focus - The robots that can learn (print edition only)

03/07/17 - Flight Safety Australia - Getting smart: artificial intelligence and aviation


28/07/17 - Technology Review - Autonome Autos: Versuch und Irrtum


12/17 - Plane & Pilot - Autopilots That Can Learn to Fly Your Plane (print edition only)


Haitham Baomar

Faculty of Engineering Sciences, Department of Computer Science, University College London,
Gower Street, London, WClE 6BT, U.K.


Copyright 2019 Haitham Baomar. All Rights Reserved.