Page 6 - Innovations in Intelligent Machines

Foreword

Innovations in Intelligent Machines is a very timely volume that takes a
fresh look on the recent attempts of instilling human-like intelligence into
computer-controlled devices. By contrast to the machine intelligence research
of the last two decades, the recent work in this area recognises explicitly the
fact that human intelligence is not purely computational but that it also has
an element of empirical validation (interaction with the environment). Also,
recent research recognises that human intelligence does not always prevent
one from making errors but it equips one with the ability to learn from mis-
takes. The latter is the basic premise for the development of the collaborative
(swarm) intelligence that demonstrates the value of the virtual experience pool
assembled from cases of successful and unsuccessful execution of a particular
algorithm.
The editors are to be complemented for their vision of designing a frame-
work within which they ask some fundamental questions about the nature
of intelligence in general and intelligent machines in particular and illustrate
answers to these questions with specific practical system implementations in
the consecutive chapters of the book.
Chapter 2 addresses the cost effectiveness of “delegating” operator’s intel-
ligence to on-board computers so as to achieve single operator control of mul-
tiple unmanned aerial vehicles (UAV). The perspective of cost effectiveness
allows one to appreciate the distinction between the optimal (algorithmic)
and the intelligent (non-algorithmic, empirical) decision-making, which nec-
essarily implies some costs. In this context the decision to use or not to use
additional human operators can be seen as the assessment of the “value” of
the human intelligence in performing a specific task.
The challenge of the development of collaborative (swarm) intelligence and
its specific application to UAV path planning over the terrain with complex
topology is addressed in Chapters 3 and 4. The authors of these chapters
propose different technical solutions based on the application of game the-
ory, negotiation techniques and neural networks but they reach the same
conclusions that the cooperative behaviour of individual UAVs, exchanging