Preface

This book is very special in several respects. On the one hand, it is the first of its kind where the reader can find in one place recent research results by leading academics from British, American, Australian and European universities as well as reports of implementation activities by leading industry-based researchers from giants like Boeing and authorities like MITRE. On the other hand, it combines topics such as human factors and regulation issues with technical aspects such as sensors, algorithms, methodologies and results. It is also unique because it reports the latest results from simulations, real experiments and implementation. Further, because the area of unmanned aircraft systems (UAS) is projected to grow exponentially in the next few decades in the most developed countries. Because of its nature (being closer to defence developments and thus being less open), publications (especially books, guides, instructions and reviews) are difficult to access. Indeed, the UAS market is forecast to grow from its present $5.9B to $11.3B annually during the next decade, totalling $94B for the period [1]. Moreover, it is envisaged that the F-35 Lightning II (Joint Strike Fighter) and respectively the Russian equivalent T-50 (PAK-FA) fifth generation jets will be the last major manned fighter aircraft types and the focus will shift to UAS. Large (multimillion) research and development programmes such as ASTRAEA, Taranis, SUAV[E], Mantis, etc. have taken place in the UK and similarly in the USA (two Grand Challenge competitions by DARPA; WASP III, Raven, Scan Eagle, MQ-9 and MQ-18, RQ-4 Blk and more recently, X47-B and RQ-170 (Sentinel) which was downed recently over Iran), leading European countries (France, Sweden, Germany, Czech Republic) and Israel during the last decade or so. UAS are critically important for future military capability in areas such as intelligence, surveillance, suppression of enemy air defence, close air support, situational awareness and missile defence. Their role in operations in Afghanistan and Libya cannot be underestimated. In 2009, the US Air Force started training more pilots to operate unmanned systems than to fly fighters and bombers [2]. The US Congress has mandated that, by 2015, one-third of ground combat vehicles will be unmanned [2].

There is also an embryonic, but very fast growing, civil market for UAS in areas as diverse and important for society as the police force, fire service, ambulance, coast guard, air sea rescue, fishing patrols, mountain rescue, utility companies, highway agencies, environmental protection, agriculture, nuclear industry, volcanoes research, postal services, communications, etc. It is reported [3] that currently there are some 300 UAS worldwide with over 100 (unsurprisingly) in the USA, followed by France and Russia and (somewhat surprisingly) the UK in 13th position with only 5, behind Switzerland, Norway, the Czech Republic, Japan and Israel.

Yet, the number of publications – and especially organised in books, guides and proceedings – on this specific topic of obvious interest is insignificant, if not non-existent. This book aims to fill the gap.

Before the reader is engulfed by technical details, it is worthwhile outlining the main topic, problem and terminology. First of all, it is important to clarify the meaning of the terms autonomy and autonomous. Broadly speaking, an autonomous system is one that can operate (including make decisions, plan actions, reach goals) without human intervention in any environmental conditions. In this sense, an autonomous system possesses a much higher level of automation and a higher level of complexity and intelligence than a (simply) automatic system, the theory (and industrial applications) of which was well developed half a century ago. In a more narrow sense, they distinguish different levels of autonomy, where the highest, sixth level is ‘full autonomy’ as described above. Below that there are five more levels starting from the lowest, first level of ‘human operated’ system, which often takes the form of a remotely operated vehicle (ROV). At this level, all the activities of the system are directly initiated by the human operator and the system has no control over the environment. The second, higher level, which can be called a ‘human assisting’ system, can perform actions if asked and authorised by the human. It can also be called ‘advice only if requested’ type of autonomy. The human asks the machine to propose actions and the human selects the actual action. At the higher, third level, which can be called ‘human delegated’, the machine suggests options to the human. The difference with the previous level is that it provides advice/suggestions even if not asked. Such a UAS can perform limited control activity on a delegated basis, for example automatic flight control, engine control. All of these, however, are being activated and deactivated by the human operator. A UAS of level four, which may be called ‘human supervised’ or ‘advise and if authorised act’, can suggest options and even propose one of them. This needs to be approved by the human operator, though, before being undertaken/activated! The penultimate level five, which can be called ‘machine backed by human’ or ‘act unless revoked’, includes UAS which can choose actions and perform them unless a human operator disapproves. This is, in fact, the highest level of autonomy of practical interest, because the highest level of ‘full autonomy’ is somewhat controversial (see, for example, Isaac Azimov's principles of robotics [4]).

In conclusion, there are several levels of autonomy and of practical interest are all levels but the last, the highest. Autonomous systems differ significantly from automatic systems known and used for over half a century. For the example of airborne systems, an automatic system would include vehicles that fly on a pre-programmed route through waypoint navigation, with landing controlled by the ground stations, payload switching on and off at predetermined points in the flight plan and capable of tracking a target. A UAS of interest (that is the subject of this book and offers huge potential for both military and civil applications) includes vehicle(s) that fly a mission based on tasks but has the ability to autonomously and adaptively react to threats and an evolving situation awareness capability, can adapt (evolve) the mission on the fly, where the payload can detect and manage the target and optimise performance, that can be activated and deactivated and where the interface between the ground and the vehicle is mission (task and information)-based, not control-based.

The topic of sense and avoid (SAA), which is also closely related to the term ‘see and avoid’ used in manned aircraft, is extremely important and was one of the main obstacles for wider application of UAS in non-segregated airspace related to the traffic safety and level of intelligence of the flying machines that are being produced and used both in military/defence and civilian domains. It has several aspects, including:

It has very intrinsic and strong links with a range of science and engineering subjects, such as:

• system engineering;
• automatic control;
• aerodynamics;
• image and video processing;
• machine learning and real-time data processing;
• decision-making;
• human–computer interaction, etc.

In this book, all of these issues are considered at some level of detail – including the implementation and experimental work which demonstrates ways to address or resolve them.

The book is composed of four parts, each one with a specific emphasis, namely Part I: Introduction (Chapters 1–3), Part II: Regulatory Issues and Human Factors (Chapters and ), Part III: Sense and Avoid Methodologies (Chapters 6–8) and, finally, Part IV: Sense and Avoid Applications (Chapters 9–11). The contributors are all experts in their field, and detailed biographies of each contributor can be found in About the Contributors at the start of the book.

An important goal of this book is to have a one-stop shop for engineers and researchers in this fast-moving and highly multi-disciplinary area, which covers many (if not all) aspects of the methodology and implementation of these new, exciting, yet challenging devices and complex artificial (yet very intelligent) systems which are bound to grow in number and complexity over the next decade and beyond. The aim was to combine the solid theoretical methodology based on a rigorous mathematical foundation, present a wide range of applications and, more importantly, provide illustrations that can be a useful guide for further research and development.

References

1. Teal Report, 2011. http://tealgroup.com/index.php?option=com_content&view=article&id=74:teal-group-predicts-worldwide-uav-market-will-total-just-over-94-billion-&catid=3&Itemid=16. Accessed on 18 July 2011.

2. L. G. Weiss. ‘Autonomous robots in the fog of war’. IEEE Spectrum, 8, 26–31, 2011.

3. UVS International. 2009/2010 UAS Yearbook, UAS: The Global Perspective, 7th edn, June 2009.

4. I. Azimov. ‘The machine that won the war’ (originally published in 1961), reprinted in I. Asimov, Robot Dreams. Victor Gollancz, London, pp. 191–197, 1989.