A major change in training engineering doctoral students has been initiated with the announcement by the Science and Engineering Research Council (SERC) of three centres to run pilot schemes of new four-year industrially oriented courses.
These new courses were recommended in a report of a panel chaired by Dr John Parnaby of Lucas Applied Technology. The panel was set up by SERC in response to widespread concern that large areas of the engineering industry wished to see a more vocationally oriented doctorate. The three centres, which have been selected from 33 proposals, are:
Each centre will recruit 10 research students a year with the first intake in October 1992.
The training that each engineering student will undertake during his or her four-year course will include the best features of the leading German and American doctoral schools. It will contain project management, industrial relevance and a number of taught courses, in addition to academic research and scholarship. As a consequence it will be necessary for the doctoral students to spend a significant proportion of their time in industry helping to solve one or more challenging engineering problems.
By the end of the period the research engineer should have developed demonstrable competences in the following areas: The appreciation of industrial engineering; project and programme management skills; teamwork and leadership skills; communication skills; the ability to apply their skills to new and unusual situations; the ability to seek the best viable solutions to multifaceted engineering problems; the evaluation of the environmental impact of industry and how to minimise it.
The extra work which this degree will involve is the reason why SERC will support the research engineer for up to four years. They will be given an enhanced basic stipend (£7000 a year plus annual increments) and a minimum industrial contribution of £2000.
In recognition of the increased costs involved in organising and managing this scheme and providing the appropriate training for the research engineers, SERC is providing a block grant of £30,000 a year to each centre plus an additional payment of £2000 a year for each registered student.
This doctorate is fundamentally different from the traditional PhD and in recognition of this the participating universities have agreed to award successful graduates of the scheme with the degree of Doctor of Engineering with the suffix EngD.
Further information can be obtained from me.
Open CACSD is the electronic newsletter of the IFAC/IEEE-CSS Working Group on Guidelines for Open Computer-Aided Control Systems Design (CACSD) Software. This newsletter is distributed approximately two or three times a month from the University of Wales in Swansea. Items covered include any subject related to the "opening up" of computer-based applications for the analysis, design, simulation and implementation of control systems. Examples of suitable topics are:
Much of the subject-area used to be part of the remit of the SERC special program, CDCTE, which funded the ECSTASY project ECSTASY was killed off by a number of factors, not least of which was the apparent superiority of commercial offerings. But much of what made ECSTASY unique and ahead of its time is still unique and ahead of its time! We believe that the important lessons that were learned are still relevant and deserve wider discussion - and, dare we say it, funding too!
Our aim is to stimulate the exchange of ideas about what an open CACSD system is or ought to be and hopefully shame the developers into giving us what we need, not what they want us to have! The first three issues contained the following articles:
A new organisation of transputer users and suppliers was launched at the Transputer Applications Symposium. TTC (The Transputer Consortium) is "the international community of transputer users and suppliers cooperating to increase their effectiveness through joint action, mutual support and sharing knowledge. Its initial funding is from participating supplier-companies and it has the non-financial support of user groups. Representatives of the first sponsor companies (Inmos, 3L, Parsys, Parsytec and Transtech) and of WoTUG (the World occam and Transputer User Group) were present as Mike Jane (TTC Coordinator) announced the launch. TIC will be run by a secretariat based at RAL and overseen by an executive board representing sponsors and affiliated user groups.
TTC's objectives include:
TTC will publish a quarterly Journal, organise an annual international conference and exhibition, promote transputer education, organise a standards forum, compile and manage consultancy and product directories, foster links with professional bodies and encourage local and national meetings. The TTC journal will absorb the current SERC/DTI Transputer Initiative Mailshot and the WoTUG Newsletter, Each issue will be about 250 pages long and will include refereed technical papers, technical articles from sponsors, an editorial, user group news, hints, tips, diary and letters sections. The first issue will be published in November 1992. Papers can be submitted at any time from now on. Guidelines can be obtained from me.
The first World Transputer Congress will be held in Aachen in September 1993. TTC will be the official sponsor of the World Transputer Congress; the merger of the highly successful Transputer Applications and Transputing series.
Membership of TTC is open to any individual or organisation with a legitimate interest in the use, development and promotion of transputers.
Over two hundred and thirty delegates recently enjoyed a two day Symposium celebrating the completion of the SERC/DTI Transputer Initiative. The five year project, funded with £3M from SERC and the DTI and with a further £1M from industrial sponsorship, ended on 31 March 1992. Entitled Transputer Applications - Progress and Prospects, the Symposium's technical programme consisted of fifteen papers finishing with a panel session. The accompanying exhibition included suppliers, publishers and the Initiative's Transputer Centres.
Rod Cook, Philip Morrow, Jon Kerridge, Tony Hey and John Soraghan reported, on behalf of their Transputer Centres, a highly diverse set of projects covering financial forecasting, vision and image analysis, adaptive signal processing, CFD tools, on-line monitoring and control, commercial parallel numerical libraries, database developments for banking and virtual channel routing. Highlights included successful registration of a Transputer Centre with the Design Council, modification of a CASE tool for use with transputers and a novel variation to a signal processing algorithm which, although developed purely to overcome hardware limitations, proved to be superior to the original.
George Irwin and Peter Undrill gave fascinating reviews of transputers in real-time control and image processing. Transputers present control engineers with the ideal opportunity to apply the computationally demanding advances made in the theory of the subject over the last ten years. Examples range from flexible high speed manufacturing machines to a flight control system successfully tested on a BAe 1-11 aircraft. Transputers allow image processing specialists to exploit computational power far in excess of that usually available to them. The huge variety of applications includes image assisted robotics, virtual reality and knowledge processing.
Ron Perrott and Peter Welch presented complementary views of the state of transputer software. The first was wide-ranging, covering languages, operating systems and environments. There remains a great need for standards in all areas. New language definition activities and the commercial sector are key areas for attention if parallel processing is to become the norm. The second view focused on occam. The importance of the language is not solely, or even chiefly, in its elegant model of parallel processing but in its security. Occam is alone in preserving the rigorous separation between evaluate and update activities required to support formal reasoning about programs. The speaker therefore raised the novel question: Can C survive beyond 1997?.
David May's address on The last 10 years and the next 10 years might well have been sub-titled What a lot of people don't know about Inmos. The development of transputers depended upon subterfuge and deception: early reticles for the profitable colour look-up chip, supplied by the Bristol development team, all included prototype transputers which the Colorado staff were therefore obliged to manufacture. Another Inmos chip design had its corners sliced off to fit better on to a silicon wafer - making it the world's first and only example of an octagonal chip!
The wide range of forthcoming activities announced during the Symposium illustrated the continuing strength of interest and activity in transputer developments. Meetings included WoTUG 15 at Aberdeen in April 1992, PACTA'92 at Barcelona in September 1992 and the first World Transputer Congress (WTC '93) at Aachen in 1993. Of especial interest was the official launch of TTC which will preserve and expand some of the work of the Transputer Initiative.
Three demonstrators developed through the Initiative's Community Clubs in real-time control and image processing were on display. HECTOR (the HExapodal CTacc demonstratOR) was put through his paces both on video and in reality. A non-destructive fault detection system examined metal tubes at speeds approaching production line rates while a signature verification system helped to identify the frauds among the delegates.
Paul Williams, the Director of RAL, was the speaker at the Conference Dinner and explained, in his after dinner speech, that we are not mourning a death but celebrating a birth. The Transputer Initiative is now five years old and its ability to survive on its own is illustrated by the ongoing success of the Transputer Centres and the launch of TTC. It ranks high in the long list of successful cooperative ventures between SERC and DTI. Colin Whitby-Strevens, a staunch supporter of the Initiative from the start, provided an Inmos perspective. Peak production of the T8 - the old transputer first sold in 1987 and soon to be followed by the T9 series - has not yet happened and will not occur until 1995. Manufacture of the (even earlier) T2 and T4 families will continue at least until the end of the century.
Many who had contributed to the success of the Initiative were thanked - notably Mike Jane without whom none of it would have been possible. Special mentions were made of two sorely-missed colleagues: Doug Lewin, who set the tone for the Initiative as Chairman of the Computing Facilities Committee but died shortly after its launch, and Barry Copestake, of the Department of Economic Development (N.I), who was fatally injured in the Kegworth air disaster.
Effective Use of Colour and Hardcopy and Video Facilities in Visualisation was the wordy title of a meeting held at the Rutherford Appleton Laboratories in mid-March. It was the first meeting organised by the Visualisation Community Club and drew together a wide spectrum of scientists and engineers for a series of talks about the production of computer output on colour screens and printers, and scientific/teaching videos.
Wide-ranging advice was offered and here are a few samples:
So, how would you do it? As a mathematician you might choose
2 1 3 4
but a clock-watcher might choose
4 1 3 2
and a book-reader
1 2 3 4
and a book-reading clock-watcher
1 2 4 3
Population Stereotypes was the second session of a 2-day course on Human Factors in Design organised by EASE at the Rutherford Appleton Laboratory. The session attempted to explain how one population of people behaves differently from another. It is all too easy for the designer of a computer interface to assume that the user will behave in a pre-conceived manner. The course then bounded on through:
The implications of these factors for user-interface design were then discussed. Guidelines ranged from straightforward advice on the presentation of visual information to the user, eg use a minimum ratio of 1.5: 1 if you have two different font sizes on the screen, through to the detailed Smith and Mosier (1984) reference tome Design Guidelines for User-System Interface Software. The latter has a PC software version called NaviTextSAM which can be used, eg to select recommended letter codes for menu selections. The history of the development of modern direct manipulation interfaces using WIMPs (Windows, Icons, Mouse/Menus and Pointers) was illustrated by a video All the Widgets. We were able to see how scroll bars have evolved over the past 10 years, showing the many different ways of accomplishing the same basic task.
Keyboards, mice, flying bats, moles, data gloves and even dead mice were some of the many input devices discussed. The course speaker had been involved in the evaluation of a future EuroWorkStation and was well-versed in the latest technology. Problems associated with voice input and auditory output were outlined, e.g. imagine asking your computer for a New Display and getting a naked couple appear on your screen Nudists Play!
The final afternoon of the course was concerned with the specification of user interfaces (informally) and User Interface Management Systems (UIMS). A couple of different systems were outlined and the problem of user interface/application separation described. One approach is to create a linkage layer of code. In the past year, because of the separation problems, there has been a trend away from UIMS to UIDEs or User Interface Device Environments. The UIDE essentially holds together a library of design tools based on objects/units.
The Peridot system developed in the US during the late 80s was shown to us on video. Essentially Peridot allows a user-interface to be created interactively without any programming. The creation of a menu system was demonstrated, in which many of the features were selected purely by computer inference of what was required. The main sessions were interspersed with several practicals and demonstrations on SUN workstations. These included:
For me, there was regular frustration as I tried to use Windows 3 techniques for a PC on a SUN workstation! Nevertheless I found it a very useful couple of days. The extensive bibliography provided at the end of the course could keep me extremely busy for a further couple of years.
(If this article has aroused your interest in the intricacies of user interface design, the Human Factors in Design Course will be run again on Wednesday, 17 June and Thursday, 18 June at RAL. If you would like to attend, please complete the Registration Form enclosed with this issue of the Newsletter.)
The use of software in safety-critical applications has grown rapidly in the last decade and continues to increase. Prominent, headline-making applications such as railway signalling, nuclear power stations, chemical plants and fly-by-wire aircraft readily come to mind, but other examples are weapon systems, hospital equipment, including life-support systems, and motor transport. Further, the use of microprocessor based general-purpose controllers as components in safety-related equipment is increasing.
The Department of Trade and Industry (DTI) and the Science and Engineering Research Council (SERC) have for some time supported the Safety-Critical Systems research programme, and together with other Government Departments and organisations - HSE, MoD, DTI, DoH, DoE, CAA and National Power - they support the coordinated approach to activities, such as standards and technology transfer, proposed in the SafeIT Initiative. Out of these programmes emerged the Safety-Critical Systems Club, which held its inaugural meeting in Manchester on l1th July, 1991 with over 250 delegates present.
Purpose The Club's purpose is to facilitate the transfer of information technology, and current and emerging practices and standards. It is a non-profit organisation set up to provide a service to all. It sets out to cooperate with all bodies involved or interested in safety-critical systems, and has no affiliation other than those described below under Organisation.
The club seeks to involve all sectors of the safety-critical community, and both technical and managerial levels within them. By doing so, it can facilitate communication among researchers, the transfer of technology from researchers to users and feedback from users, and the communication of experience between users. The benefits will be more effective research, a more rapid and effective transfer and use of technology, the identification of best practice and the definition of requirements for education and training.
Responsibility for running the Club has been contracted by the DTI and SERC to the British Computer Society (BCS) and the Institution of Electrical Engineers (IEE) jointly. The Centre for Software Reliability (CSR) has been contracted to organise the club. The CSR has accumulated considerable experience in this field through running the CSR Club in Software Reliability and Metrics since 1982.
The principal means of achieving the Club's goal of communication is by holding at least four events and publishing three issues of a newsletter per year.
The newsletter is free to members and offers editorial, articles, an event agenda and other information useful and interesting to the safety-critical community. Contributions are welcomed and should be sent to Felix Redmill, SCS Club Co-ordinator.
The events will be of one or two days duration in the format of workshops, seminars or conferences, as appropriate. There is a modest attendance charge. Events offer the opportunity to meet others with common interests and to share experiences, to listen to experts speaking on their research projects and their experience of the use of technology, and to participate in both formal and informal discussion. Members receive advance information of all events.
There is also an annual general meeting, normally arranged to take place immediately before or after one of the events. In addition, special events or mailings are arranged for particular sectors of membership, when appropriate.
Membership is open to everyone, and there is at present no membership fee. However, as it is in the Club's terms of reference eventually to become financially self-supporting, this may be reviewed later. Members receive, free of charge, all issues of the newsletter, information on forthcoming events, and any special interest mail shots. Further, by participating in Club activities, they can help to determine the Club's direction.
The expanding UK research community in the area of Engineering Design now has its own electronic bulletin board called ENGINEERING-DESIGN. The new bulletin board has been set up on the SERC supported list-server called Mailbase developed at the University of Newcastle upon Tyne.
The Bulletin Board is an electronic mailing list serving the researchers at the five Engineering Design Centres and other research groups active in engineering design. The membership is currently open to all. Monthly archives of ENGINEERING-DESIGN are available for occasional readers.
Topics to discuss can vary as far as your imagination stretches within the realms of engineering design and the technology that affects it. One could even start a meta-level discussion on engineering design itself.
The list encompasses, among other things:
To Join To join the list ENGINEERING-DESIGN, e-mail the following line:
subscribe engineering-design your_firstname your_surname
to: mailbase@uk.ac.mailbase
The theme of the conference is the potential of new Advanced Database Systems. Relational systems are now able to run on a network of Parallel Processors. Object-Oriented systems are being developed which directly support new object-oriented systems analysis techniques. Novel types of databases are being proto typed which allow the storage of code in the database using new styles of programming for constraint checking and greater productivity: Functional, Persistent and Logic Programming. This is of particular interest to those working on CAD or engineering design. Papers will be presented on all these topics.
The invited speakers are Michael Brodie (GTE, Boston) who will address the serious problem of Legacy Systems and moving older applications onto such advanced systems. David Gradwell (Data Dictionary Systems) will address the issues of Object-Oriented Requirements Capture and Analysis. The conference is a meeting place for those interested in academic research and industrial take-up. It shows a vigorous database research community willing to address issues of industrial importance.
The conference takes place on the beautiful old campus of Aberdeen University, in halls looking across lawns to the 16th century King's College Chapel. The conference dinner is in Haddo House (courtesy of the National Trust). There will be an exhibition and demonstration showing software developed in Scotland, which has many university database groups.
