The development of new surgical techniques such as Minimally Invasive Surgery (MIS) and the recent interest in selection and assessment of surgical trainees has high- lighted the issues that contribute to the acquisition and development of the psychometric skills needed by a surgeon. This presentation will identify the potential and challenges associated with providing an effective Virtual Environment MIS training system and will describe the VE-KATS system developed at Hull to assist trainee orthopaedic surgeons acquire and enhance the skills required to perform arthroscopic surgery.
In addition to its use as a training system, the presentation will highlight the use of VE-KATS in assessing the utility of new (virtual) surgical instruments such as stereoscopic arthroscopes without the need to produce the physical artefact.
The advent of a National Lottery in the United Kingdom prompted the thought about how one would design a highly parallel system to manage such an activity, especially, as the solution adopted by the lottery company was to use a parallel database machine. The fact that the selected machine does not use a database management system but uses bespoke software makes the exercise even more interesting. The aim of this design exercise is to investigate how a highly parallel scalable computer system could be constructed using a transputer based architecture.
The field of automated theorem proving witnesses yet another shift of its research paradigm: back to its (AI)roots.
This talk will make this change of focus explicit by giving a historical perspective in the first part of the presentation.
In the second part I will give an overview of an automated theorem proving assistent currently under development at the University (DFKI) of Saarbruecken.
The talk begins by surveying some recent results in computer science which suggest that computers might indeed become superior to human beings. These include a computer capable of defeating the (human) world chess champion, and a machine learning program which has discovered previously unknown laws concerned with protein folding. Consideration is next given to the argument developed by Lucas and Penrose to the effect that G del's incompleteness theorems show that the human mind can transcend any possible computer. This argument is rejected in its original form, but it is claimed that a variant on the argument can yield some interesting results.
The Centre for Music Technology at the University of Glasgow is an interdisciplinary unit involving the Departments of Music, Computing Science and Electrical Engineering. The systems developed and in use to date have concentrated on single user systems for composition, teaching and performance.
The Centre has just started several projects to investigate the issues arising in a distributed music environment: MINIMS is both a scientific investigation into the degree of synchonization demanded by professional musicians and a technology programme to develop the systems to achieve this goal; in parallel NetMuse is deploying an infrastructure between seven HE establishments within Scotland, initially with the goal of real time access to music archives and performance broadcast, but planning to migrate to supporting distributed interactive sessions building on the MINIMs work.
The talk will present the challenges the project is facing, and they are scientific, political and educational as well as technological.
This talk eill offer an assessment of what has been achieved in three decades of work on the semantics of programming languages and pinpoint some practical problems in computing which might stimulate further research. The main reason for doing this is to attempt to persuade leading researchers in our field to encourage some of their younger colleagues to tackle practical problems with theories which are available. All too often it seems that publications are aimed at devising refinements of theories which --whilst they appear to derive their interest from computing-- may not be applicable to a useful class of applications. The examples sketched come from the author's own work on concurrent object oriented languages, from database work, and from more speculative research.
In modern informatic systems, computing and interaction are inextricably intertwined. Yet several dichotomies separate these two notions as they are normally understood:
COMPUTING: sequentiality, determinacy, prescription (of evaluation)
INTERACTION: concurrency, nondeterminacy, description (of interactive behaviour)
It is a serious challenge to find a semantic theory which unites the two.
For this, we have to go deeper than the ideas found in concurrent or object-oriented languages, which usually comprise familiar sequential programming notions with a top-dressing of communication.
In this lecture, I propose that naming is the key to a unifying theory. I support the thesis by giving simple examples in the Pi calculus, which was introduced as a calculus of mobile processes in around 1990 by Milner, Parrow and Walker. In the Pi calculus names do multiple duty - they may be addresses, pointers, references, channels, handles, variables, ...
This allows one not only to express mobile systems but also to derive both imperative and declarative constructs of familiar programming, including the Lambda calculus. My title derives from the following alignment:
COMPUTING: a name denotes (a value)
INTERACTION: a name addresses (a process)
What is the use of such a unifying theory?
Students in the second year of the Computer Science Tripos work on practical group projects during the Lent Term. This seminar will consist of a series of brief presentations by representatives of some of this year's groups. Their topics include:
There will be an opportunity to see some of the work being demonstrated before the seminar between 3pm and 4pm in the Cockcroft 4 Laboratory.