Convenor's Report of Proceedings

"To hell and back, yet again - or are we still there?"

This report is formed from notes made by the Convenor, Peter Sydenham plus presentations by invited speakers plus surveys formed by those present.

The Academic Forum seemed to be a success, but this is what you would expect the Convenor to report! (The 23 Feedback sheets showed ‘Best Yet’- 34%; ‘Cannot judge’ – 56%; ‘No’ - 10%, and with 2 suggesting the discussion started fine but tailed away into old, stale issues.) It was, for the first time of a Forum/Workshop, organised with the usually required elements of an event of this nature.

Other reports and supporting information are introduced here where appropriate as hyperlinked documents all operating within the same folder.

It began with most of the 38 delegates arriving together at the venue in the Department of Computer Science and Software Engineering, University of Melbourne in a bus from the Conference Hotel. The group consisted of around one third being those who consistently attend and/or organise Forums so there was immediate warmth in the relationships in evidence. The assembly represented academics and industry people from countries including the US, UK, Israel, Norway, France, Germany and Australia.

The program was used much as it was originally set up. The main schedule change en-route was that the presentations using a little more time than budgeted had their follow-on debate times rolled into the last debate.

The venue was excellent being a modern lecture theatre with electronic projection, lecturer room control console and good climate control. All presentations were in Power Point and these went mostly to plan expect for one file that opened with some missing slides.

The proceedings went along with vigour - to say the least. There were no dead patches and many attendees probably felt they still wanted to contribute to the debate to the very end. The flow of issues strayed a bit at times but generally tried to come back to the theme at hand - Advancement of Systems Engineering through Academic Scholarship.

Delegates were issued with a glossy covered booklet giving the Convenor’s Foreword, the Theme statement, program and resumes of the presenters. They also received in addition the latest program, an issues sheet for recording points they felt were worth recording for later attention, a delegate list and a Forum feedback sheet.

Those present represented a reasonable balance between academia and industry thinking. They were certainly not all of the same kind of ‘conversion’ but could be said to have been thinking along the same lines as systems engineers. (PHS note: in organisational science terms it was not a ‘singular’ group but certainly reasonably ‘coalition’, and certainly not ‘coercive’). Some useful inward looking points were made by the host delegate, Leon Sterling, who felt he was looking into a somewhat different world than his own of software engineering.

This report will not attempt to give a blow-by-blow account but pick up on some highlights and be used to integrate the various aspects of the event. The topics covered were many and there was not always agreement.

The debates still continue! It is expected that the reader will view the appropriate presentations as they use this report. Some felt we had returned to very much old ground but others clearly felt it was productive.

The Forum started on time being opened with a pleasant welcome by the Senior Deputy Vice Chancellor (Senior VP) Professor Sally Walker of the University of Melbourne.

The debate that followed Donna Rhode's scene setting session, in which she first set the theme scene by reiterating the ‘Advancement’ issue of the INCOSE strategic plan, rapidly got round to the place of academics in the SE scheme of things - a little earlier than desired by the staged speaker order.

Much of this session’s following debate was somewhat removed from the theme of INCOSE needs, per se. It got through many issues and had to be finally guillotined to meet the program break. Some points from the ensuing strong debate were:-

• Academic SE does not have a place of honour and respectability in the campus community

• Strong academic leadership is needed to keep it going as a robust critical mass that can support its arguments in the collegiate environment. INCOSE

• Multidisciplinary approach is essential.

• INCOSE could help the above by lobbying for university support in key funding bodies.

• The distinction between the Breadth and Depth conflict and its parallel academic vis pragmatic conflict were a constant metaphor that is universally understood as a good simple descriptor of the educational challenge. The T diagram was stated (Which, by the way seems to stem from software V&V author Jeff Grady, who cites its source as a summary remark from a military colleague, during the Vietnam conflict, who summarised Grady’s overview of SE with the T concept!)

• The aged old statement that better industry/academic collaboration is needed was raised (but is this a real issue or a nice throw away perception by those not familiar enough with more than an ex-undergrad’s limited view of campus operations!)

• INCOSE could organise a data base of research grant opportunities in SE

• A reference curriculum, or set of such, was needed urgently

• The political element (campus, government, national) exists as a barrier to progress in the uptake of SE as the suppliers of practitioner systems engineers (PSE’s) tends to be pulled on by the $ than by principles that used to be associated with scholarly thinking about how to supply people with long lasting education.

• A pragmatic way to develop better SE in academia is simply to get on with it and spend less time debating its foundations. Mike Woodhead gave examples of this and it seems to be true that most current new developments were of this form. Evolution not revolution! MIT still provide SE education as part of several appropriate programs and does not yet have a unit operating under the SE name. Loughborough Uni, UK is developing along these lines until it reaches the critical mass and size that will allow a centre of excellence to be declared, or even a full blown academic department/school.

• Government research and special teaching funds go where industry needs them.

• Grow where you can, in any suitable clothing, as long as excellence is retained, seems to be the maxim for the majority of campus developments supporting SE.

• Create a centre for SE by identifying the appropriate breadth disciplines and add in depth disciplines to suit the profile of the demand. Do not try to collect both B and D elements into one entity as that produces yet another kind of stovepipe (but can one ever manage efficiently without a stove-pipe organization?) Must balance the B and D and maintain excellence in both.

• Team based, project based teaching is essential with industry and face-to-face academic staff involvement (but surely this has been standard engineering educational practice for decades?)

• Surveys of courses and programs are badly needed and much overdue and how courses integrate into the B of SE.

• SE practice, and thus education/training, is too much related to defence

• Smaller countries that the US may need exposure to the very existence of SE as a body of knowledge and thinking.

 

After morning refreshment, provided by the host school, the next session opened with a presentation by Phillip M’Pherson, the person possibly with the longest experience in academic SE leadership and application. He put into excellent perspective the issues that need to be considered. Some points were:-

• ‘SE is like a virus that gets into everything’.

• Some say that academics teach subjects as ‘closed’ systems in which the boundaries are selected to leave out the difficult bits.

• ‘SE is what engineering aught to be’

• Set up the cross bar of the T in the undergraduate course with one or two D’s hanging under that. Add D’s as the PSE matures to eventually be a block.

• Need to better define the audience and not frighten others away.

• The legacy problem of knowledge is a critical issue. One INCOSE role is certainly to assist maintain the SE body of knowledge. Academic role is to generate, sort and disseminate knowledge to people. (PHS note: look up your local University’s charter, study the history of the ‘campus’ from its Cairo days; view that enthralling film ‘The Name of the Rose’ with Sean Connery.)

• ‘If people understood the value of SE all would want it’. (PHS: See Honour paper 7.3.4 in IS2001 proceedings)

• The presentation had gems of wisdom and seems to be compulsory reading (– once the file is corrected.)

• Regarding the transference of skills and knowledge to new generations of PSE ‘for how long’ and at ‘what level’ are issues to consider

 

The Panellists then gave their presentations. Yupo Chan gave an outline of the circumstances and nature and scope of his new SE program. Leon Sterling, the host of the Forum and keen to see how his field might benefit from SE, brought in the views of a ‘close relative’ - software engineering. He gave comment on how the meeting was tackling its contemplation issues. Andy Sage gave us the benefits of his long activity in SE education program delivery and what that has distilled out for him regarding the nature and structure of SE. Stephen Cook covered the developments of his University and how they support stakeholders.

The planned debate on these papers was deferred to later and all went to lunch in the foyer.

After a good lunch of hot finger food and sandwiches Bob Tufts ‘reversed the flow’ giving his views on what industry expects from academia. Using the experiences of his large employer (PHS comment: it is not uncommon for some companies to have over 1500 PSEs on staff!) He concentrated on the staff hiring aspect. Some key points were:-

• Is there any commonality in the nature and content of SE courses and awards?

• Industry can’t wait for universities to ramp up a response as they have a huge backlog of unfilled SE posts – Bob mentioned figures in the hundreds in major systems companies

• This has resulted in the large corporations setting up their own University/Institutes – Grumman is likely to be next. These are usually done in close association with cooperative universities

• Only a low percentage of invited universities are willing to join such coop ventures

• Fast track education is essential for companies needs – but how much of their need is really education (basic, long term, principles for supporting reflective practitioners) or training (application of practice to current rules and norms with short term benefits)?

• Bob’s presentation spells it out.

This session began by breaking out into many questions - not all agreed with his points. (PHS comment: For example, have corporations developed their own teaching organizations as a means to dictate the content of SE and keep it more proprietary as their crown jewels - or is this truly a reaction to the stated lack of campus interests?)

In the interests of hearing the views of the many it was cut and the industrial panellists presented. David Wright gave an overview of SE but rapidly got into the need for a scientific approach to identify what to teach (PHS note: Does he mean ‘scholarly’ instead of ‘scientific’ for not all human intellectual thinking is accepted as being ‘scientific’?). He clearly pushed for a: -

• Crisper approach to finding a sound solution to the call for sounder SE education.

David later submitted a short post Forum report on his thoughts. The content of that report also relates to comment about the final debate given later in this report.

Mary Simpson, Battelle, USA gave a concise overview of what she expects from the effects of academics on students. She made the case for

• not giving so much effort to finding new knowledge but to educating people to better use the vast amount of already existing knowledge.

Jean-Philippe Lerat

contributed insight into the work of the ISO group he is with who are seeking to standardise mature SE processes. They suggest the key capability areas are ‘technology’, ‘fundamental knowledge’ and ‘skills in sound communication’. He suggested that much more emphasis is needed on the communications aspect, as this is the least understood one. It is not just a matter of passing messages (through interfaces) but also in understanding them. A key point he made was:-

• the PSE carries major responsibility for the quality and adequacy of communication between all humans involved in the execution of a program.

The issue of ‘closed’ versus ‘open’ thinking came up a lot and it was considered we need to take more of an open thinking attitude and educate people to be able to do this better.

To end the pumping up of input on expressions of ideas on the theme Dennis Buede gave a comprehensive presentation on the issues of teaching SE and how the practical aspect is dealt with at his University by the use of Lego robots to bring out the concepts of SE. His presentation included many very apt quotations.

This got the meeting to afternoon refreshment time after which the final debate of the day took place. It was expected that some key issues would be identified and debated -- but we were to be disappointed.

Errol Lawson had agreed to moderate the final hour of debate which was to explore how to bring out the salient issues and points for action. He began with a presentation that started by simply quoting the existence of Kline’s book that laid down the steps in which a multi-discipline develops. He then put up the first 4 steps of Kline’s eight.

The question was put to the delegates ‘could the assembly agree with these steps, starting with Step 1. Selection of a Class of Systems with and Associated Set of Problems’?

For a while the debate generally agreed with the first step but that rapidly fell apart as the words were debated with a vocal few not agreeing with some. The protagonists of the ‘no’ case were challenged to put up another hypothesis for deciding if SE is a discipline but an alternative hypothesis did not come.

Perhaps the delegates had become dulled by this time of day. Some rallied for the ‘yes’ case and made the point that there were those present who were viewing SE in a pragmatic manner rather than looking for more scientific underpinnings. David Wright’s later report, mentioned earlier is relevant to the ‘yes’ case.

It seemed to some that the Kline structure is trying to forced the reductionism viewpoint when the problem may well be more of the soft kind requiring the phenomenological approach - or does this problem need the latter followed by the other thus using both forms of thinking? Some points raised were:-

• Where does SE add value?

• What is the belief system to use, or being used?

• Do Kline’s steps for the date period to 1900 apply for post 1900 developments? Do we think differently today compared with then?

• Mature disciplines do not seem to define themselves but get defined by others as they create emergent properties. They just ‘do’ by getting on with it until they emerge and are accepted

• SE is a success story to a degree but it must be packaged for sale

• Let’s stop this contemplation and self-flagellation and cease seeking a deeper understanding is one view. Others say this is not the way for that is just what SE has been doing via numerous experiments – i.e. the many real SE projects of the past

• Formal ‘rules’ for operating a program’s engineering of systems are often declared to be the company crown jewels. INCOSE could seek make these more available as that would help advancements by removing a potential barrier to progress

• So much of the activity of being an PSE is the application of tacit knowledge - the stuff you do not know you know until you apply it to solve problems then realising that less experiences and less capable people do not have it

The discussion was again halted – after a whole day of this there were still those present who could have gone on debating! One delegate was compelled to make a statement that whilst he was usually reticent to speak at meetings he had been incited to burst forth!

This meeting is actually just one of three similar events that have taken place over the past year to set in place some recorded discussion on the foundational academic issues of SE. They each give views of different groups of people on the issues to be faced in supporting academic SE, the SE constituency and other benefactors at large.

These meetings were the:-

• ‘Academic Morning’ held as part of the Autumn Assembly of the UK INCOSE Chapter www.incose.org/uk/ November 2000 (a similar invited presentation event convened by Peter Sydenham)

• ‘Seminar on SE’ organised by Norwegian NORSEC INCOSE Chapter www.incose.org/norsec/, June, 2001 (organised by Cecilia Haskins to bring the SE academics in Norway into conjunction with ideas from the US, UK, Netherlands and Australia)

To end the day Peter Sydenham presented all invited speakers and chairs with IS2001 appreciation certificates. Somehow this seemed to not totally acknowledge the contributions made for the delegated all entered into the debate in a serious manner making it an ‘all persons equal’ event.

No serious student of this day’s theme should miss reading the papers of the Special issue on SE published before many PSEs of today were born! See Engineering Education Journal, April 1970. It makes interesting reading for much of that material and reviewed situations have not moved on for 3 decades!

The paper ‘Systems Engineering: The Ubiquitous Unicorn?’ by John Davies, that was recently presented in the UK Spring INCOSE meeting, is also relevant. This was handed in to go with this record.

Peter Sydenham has published tutorial statements on this theme in the IS1999 symposium, at the 1999 UK Autumn Assembly, and in a 1999 issue of INSIGHT (the INCOSE newletter/magazine).

Some Conclusions.

Where did the day take us? Was it worth our effort?

• Possibly the main contribution is one that may seem trivial to some but was badly overdue. The event set the bar at the expected right level regarding the format and delivery of an Academic Forum. Previous events lacked academic organisational rigour. Nothing remarkable was really done here – just proper attention to detail in all aspects as would be expected for a one-day seminar. With this baseline set continuous improvement can now take place within future symposiums etc.

• Next, we collectively have here set down a record of a set of discussions that others can use to temper the repetitious debates and navel contemplation that has been seen in previous forums – and no doubt in tearooms etc. around the world. Record taking has not happened for previous forums so we had to rerun the game plan and play to get data to work up from.

• Many key issues were identified that may be found useful to others. Just how they should best be tackled is not clear. They impact on several technical groups of INCOSE so need to be considered by more than another Academic Forum. Perhaps it makes sense to now regard the Academic Forum as an interest group activity within the INCOSE technical community?

We did not resolve the debate but we got more insight into its variables. The bogey of what SE is still hangs around? Are we reinforcing it by debating it to death? Is it really a matter of simply letting it emerge from doing it in an ever better way?

The usual event comments sheet on the Forum was issued. 23 were returned. An analysis is found by studying the summary. These data should be useful to future organisers of Forums.

Additionally an Issues Record sheet was provided. Delegates were invited to record key issues as the day passed. Sixteen returns were received with considerable comment upon them. These have been assembled into an issues compilation without any editing. They make interesting reading and give a record of the fundamental approach taken, and the thinking behind them, of a diverse set of practitioners including many who were not academics or also were both academics and industry folks over their lives. They show the need to keep addressing this theme but not as a talk-fest without clear, bounded requirements.

It was a pleasant project to undertake. Those who supported the day in the many ways needed for success are thanked as it was a joint effort to get it together and running as well as it did.

Peter Sydenham

Adelaide 7 July 2001

Sydenham@senet.com.au