|
|
 |
|
|
Real Artificial Intelligence - Speakers
Full details of the speakers and their presentations are given below.
|
|
|
|
Professor Adrian Hopgood
Dean of the Faculty of Technology, De Montfort University
Artificial Intelligence Demystified
Artificial intelligence has been a rich branch of research for computer scientists and psychologists for over 50 years. The concept of mimicking human intelligence in a computer fuels the public imagination and has led to countless academic papers, news articles, and fictional works. Such exposure has led to high public expectations, despite the incredible complexity of everyday human behaviour and the difficulties in replicating even limited aspects of it.
The challenge now is to build a system that can operate across the spectrum of intelligent behaviour from low-level reaction and control to high-level specialist expertise. The achievement of this goal requires a hybrid approach that draws on a variety of different techniques. Several practical examples will be presented, ranging from the control of specialised manufacturing processes to the diagnosis of mouth cancer. No prior knowledge of artificial intelligence will be assumed.
Professor Adrian Hopgood joined De Montfort University in 2007 as Dean of the Faculty of Technology, having previously worked for Nottingham Trent University and the Open University. He also has industrial experience with Telstra Research Laboratories in Melbourne, Australia and Systems Designers plc (now part of HP)
Adrian has published widely and his text book “Intelligent Systems for Engineers & Scientists” is ranked as a bestseller. He is a visiting professor at the Open University, Fellow of the British Computer Society, Chartered Engineer, and a panellist for the Engineering and Physical Sciences Research Council (EPSRC). He holds a doctorate from the University of Oxford and a bachelor’s degree from the University of Bristol.
|
|
|
|
Colin Cadas and Badr Haque
Rolls-Royce
Whatever Happened to Expert Systems?
AI-2009 Keynote Application Address
In 1989, the expert systems community knew there was too much hype: it was well publicised. We just didn't know “how much too much”. Yet even today, we look back and say “Whatever happened to Expert Systems?” – as if there was some sort of golden age and things have never been as good since – the way old people remember the weather.
So what is the truth? What really happened in the 1980s with expert systems? What went wrong and why? Has anything happened to “fix” expert systems since? What is the situation today? What will happen in the future? These are the questions we all attempt to answer. Colin will give a historical look back from the point of view of industrial implementation, citing examples from within Rolls-Royce and outside, finishing with a look at the current state of the art within Rolls-Royce and a look ahead to the future. Badr will complement this view with a look at a few real applications deployed within Rolls-Royce, discussing how they were developed, how they are used and what the benefits are.
Colin Cadas graduated from Cambridge University in 1985 in Engineering. He is a chartered engineer and a member of the British Computer Society. Colin started his career 20 years ago in Rolls-Royce as a methods engineer developing nuclear containment for the company’s submarine business. Since then, he has worked extensively in knowledge acquisition, knowledge modelling, expert systems, real-time systems, artificial intelligence, software engineering, knowledge management, semantic web, and design engineering processes.
He is currently “company specialist – Knowledge Management” in Rolls-Royce. In this role, he is responsible for both the research and technology acquisition strategy in engineering knowledge management, and the knowledge management implementation and support strategy. This involves the acquisition, development and implementation of knowledge management techniques and practices across the company, working with all business units within R-R to implement the best tools and techniques. Though the work has an engineering focus - the tools and techniques are deployed across the global Rolls-Royce engineering community (9000 people) and beyond. His personal interests include music, sport and travel. As a lover of good food, Colin hates mayonnaise and other mayonnaise-like foods.
Badr Haque is currently manager & technical leader for bespoke “Design Systems” within the Fans Supply Chain Unit of Rolls-Royce Gas Turbine Operations. These systems improve the engineering or ‘virtual product development’ processes through the use of automation and integration technologies including Knowledge Technologies.
Badr graduated with degree in Aeronautical Engineering (1991), followed by a Masters in Flight Dynamics (1992). Following a brief period in industry in various roles including mechanical engineering and manufacturing planning, he joined Nottingham University (1995) as a Research Associate. His research focussed on the application of Information and Communication Technologies to improve Concurrent Engineering and Systems Engineering processes. He worked on five European research programmes in various manufacturing industry sectors. He gained experience in design, development and application of business process modelling technologies and knowledge based systems, as well as the application of object oriented modelling and internet technologies. After receiving his Doctorate (1999) he joined the DTI/SBAC ‘UK Lean Aerospace Initiative’, where he led the development of methods & tools for Lean Product Development. In 2001 he joined Rolls-Royce as a specialist in Knowledge Based Engineering systems, and has successfully delivered a number of KBE applications for use on Gas Turbine Engine projects.
|
|
|
|
Ashraf Elsayed
University of Liverpool
Corpus Callosum MR Image Classification
AI-2009 Best Refereed Application Paper
An approach to classifying Magnetic Resonance (MR) image data is described. The specific application is the classification of MRI scan data according to the nature of the corpus callosum, however the approach has more general applicability. A variation of the “spectral segmentation with multi-scale graph decomposition” mechanism is introduced. The result of the segmentation is stored in a quad-tree data structure to which a weighted variation (also developed by the authors)
of the gSpan algorithm is applied to identify frequent sub-trees. As a result the images are expressed as a set frequent sub-trees. There may be a great many of these and thus a decision tree based feature reduction technique is applied before classification takes place. The results show that the proposed approach performs both efficiently and effectively, obtaining a classification accuracy of over 95% in the case of the given application.
|
|
|
|
Chris Waldron
Antycip Simulation Ltd
AI in Simulation and Gaming
Chris Waldron of Antycip Simulation will give us an overview of the development of AI techniques widely used in simulation and gaming, supported by practical examples from Antycip’s products from the domain of military simulation and other spheres.
|
|
|
|
Francesco Maurelli
Heriot-Watt University
Underwater Robotics: from a Student Competition to Real Applications
When the autonomous underwater vehicle Nessie IV started its final run during SAUC-E 2009, the public was numerous on the sides of the big tank. A BBC news team was present to report on the event. Why so much interest for a student competition? In reality, SAUC-E is much more than simply a competition between teams of students. While it aims to encourage new ideas and concepts in the underwater domain, it presents real challenges to be faced and opens the way to much more complex tasks and real applications. This talk will present the key features of the robot and our approach to the tasks of the competition, but with a clear vision to the future (and, in some cases, already to the present). The missions performed are not just in the frame of SAUC-E, and differences and similarities to industrial and military applications will be highlighted.
Brief overview of the competition and of Nessie IV robot
The Student Autonomous Underwater Challenge – Europe (SAUC-E) is a competition for European students to foster the research and development of underwater technology. Sponsored by EPSRC, MOD, DGA and industrial partners, the challenge started in 2006 and is now in its fourth iteration. Eight contenders from three different countries (United Kingdom, France, Germany) competed last July in Gosport (UK), in order to demonstrate high level robot capabilities and autonomy. The tasks needed to be completed autonomously included mid-water moving target tracking, detection and hovering over a ground target, wall inspection, gate avoidance and a docking scenario.
 Nessie IV is the entry from Heriot-Watt University (UK). Built on top of Nessie III, winner of the SAUC-E 2008, this robot is a robust and reliable research platform. Its sensory system includes four underwater cameras (of which two mounted in a forward-looking stereo vision system), a 360 degree scanning sonar, depth/altimeter/temperature sensors. With its software modularity and good performance shown in trials and during the competition, it has become a key experimental platform for the Ocean Systems Laboratory at Heriot-Watt University..
Francesco Maurelli is a Research Associate and PhD candidate at the Ocean Systems Laboratory, Heriot-Watt University (Scotland, UK).
Following the Master program in Robotics and Artificial Intelligence at Sapienza, University of Rome (Italy), he prepared his thesis at Fraunhofer Institut IAIS (Sankt Augustin, Germany). His work was focused on a novel 3D laser scanner for autonomous vehicle navigation, used also in the framework of the Darpa Urban Challenge, with Team Berlin. He joined the European Marie Curie Research Network FREEsubNET in October 2007, at Heriot-Watt University. His work is focused on two main parts: standards for autonomous docking (manager of the work package), and AUV (Autonomous Underwater Vehicle) navigation, with an emphasis on sonar-based localisation and SLAM, using probabilistic techniques. He is Team Leader of Nessie Team, which has won the SAUC-E 2009 (Student Autonomous Underwater Competition – Europe). He is author or co-author of more than ten publications in international conferences.
|
|
|
|
Dr Rick Adderley
A E Solution Ltd
Fighting Crime with AI
The chances are, you or someone you know has been a victim of crime, it affects many of us during our daily lives. We see it every day in our news papers and TV news programs, crimes ranging from those that are fairly common such as burglary, robbery to the more major crimes such as rape and murder. Who deals with these… the Police. As members of the public we expect a certain level of service from the Police, however, that service has to be provided within the constraints of the economic climate and lack of resources. All Forces are affected, having to reduce staff, overheads and operating costs whilst still providing an acceptable level of service.
Applying (semi) automated data mining techniques on the vast amounts of disparate data that is captured as a result of natural day-to-day transactions, it is possible to provide operational intelligence at the “coal face” of policing and enhancing the current provision of intelligence using the same or even less personnel. Here lies another challenge; how to get data mining accepted by those same personnel; analysts, detectives and front line Officers.
Dr Rick Adderley is a retired Police Inspector having completed 32 years service. He has been involved in all aspects of policing ranging from major investigations as a detective through operational traffic duties to uniform patrol duties. His final role was the development, implementation and training of a data warehouse intelligence system that has been taken up by all 43 UK Police Forces. Rick has a Ph.D. in Computer Science (Data Mining), a 1st class honours degree in Computer Science and is a Chartered IT Professional within the British Computer Society. He has been a member of the Natural Computing Application Forum since 1999 and was one of the organising committee members for five years.
A E Solutions (BI) are a small company formed in 2003, staffed by former Police personnel, specialising in data mining, neural networks, rule induction and decision trees. The majority of the company’s focus is in the UK policing sector using their data mining workbench tool, Authority Miner, to model criminal behaviour and real world environments. Due to the nature of policing, the company has a flexible approach and the ability to dynamically respond to current issues having recent relevant experience of EU Projects under FP6 involving the zero distraction of Police motorcyclists (MoveON) and the tracking of terrorists movements (i-TRACS) and FP7, training managers in dealing with major incidents such as an aeroplane crash, terrorist bomb etc. (CRISIS).
|
|
|
|
Simon Overell, True Knowledge Ltd
Using AI to get Answers from the Internet
True Knowledge is a pioneer in a new class of Internet search technology that's aimed at dramatically improving the experience of finding known facts on the Web. Their first service - the True Knowledge Answer Engine - is a major step toward fulfilling a longstanding Internet industry goal: providing consumers with instant answers to complex questions, with a single click.
Picking up where search engines leave off, True Knowledge's path-breaking Answer Engine automates the laborious, time-consuming work that users generally must do to get final answers to their questions. True Knowledge does this by structuring data in a way that enables computers to work and think like humans do, drawing inferences and conclusions when needed to find the information that's requested. Another key differentiator: True Knowledge is tapping subject matter experts around the globe to build its information repository - bringing together the benefits of machine-driven automation and people-driven intelligence.
Simon Overell of True Knowledge will lead us through the story of how they applied AI techniques to make the break from search engines that give links to search engines that gives facts.
Simon Overell joined True Knowledge's data mining team at the beginning of 2009 where he has worked on extracting data from Wikipedia, Entity Resolution and lead the Web Mining and Natural Language Processing projects. Before coming to True Knowledge, Simon worked for Yahoo! Research Barcelona on supervised classification of Wikipedia articles and Flickr tags. Simon completed a PhD at Imperial College London in Geographic Information Retrieval and a Master's in Artificial Intelligence; he has published papers on how people use place names in context, place name disambiguation, classifying articles and geographic information retrieval. He has been a committee member of the BCS Information Retrieval Specialist Group since 2006.
|
|
