Interview with Professor Brian Collins CB

Professor Brian Collins was born in Orpington, Kent (now in the London borough of Bromley), in 1945.  His father was a quantity surveyor for Middlesex County Council and his mother stopped working to look after Brian and his older brother.

Brian went to a local primary school in Orpington. On passing the eleven-plus he went to Chislehurst and Sidcup Grammar School, based in Sidcup, which he describes as “one of the best grammar schools in the neighbourhood.  In fact, by the time I left it was one of the best in the country, vying with Manchester Grammar School.”

 Being academically orientated, Brian enjoyed school. As well as his studies, he also enjoyed playing cricket and was involved with amateur dramatics, he says: “I wasn’t on stage, I was helping people who were on stage and it rubbed off on me after a bit that I actually understood the concept of performance. That has stood me in really good stead in a number of places latterly in my life, when I understood that delivery of a message concisely was really important.”

Brian chose a science route through school studying maths and physics, although he says that in retrospect, he would have liked to have had the opportunity to do more geography and history. Having achieved seven O levels, he went on to excel at A levels and with an open scholarship, Brian went to study physics at St Peter’s College, Oxford in 1964.

 Before going to Oxford, Brian completed a six-week course at the Royal Armament Research and Development Establishment, RARDE, just north of Sevenoaks. It was here, there Brian had his first encounter with both analogue and digital computers. He says: “I did some mathematical calculations and then some programming, which to begin with they wouldn’t tell me what it was about, but I knew enough physics to know that it was to do with radiation and heat transfer.” As he was leaving, he asked what the project was and was told it was the rate of expansion of a hydrogen bomb fireball.  Brian continues: “At the age of eighteen in the 1960s, given everything else that was going on, (the Bay of Pigs, the escalation of the Cold War), it rather brought it home to me that these times and distances were measured in tens of miles and I lived ten miles out of the centre of London and you’d got about a hundred milliseconds before you didn’t know anything.  That was rather sobering.”

 Having gained his BSc in Physics, Brian continued his studies between 1967 and 1970 to gain a MA DPhil in Astrophysics, during which he used a PDP-8 computer, only the second one in the country.  He says: “It was a real privilege to have what was a ridiculously primitive machine by modern standards, 8k of memory and a 12-bit instruction set. I learnt how to program very, very efficiently with that. The engineering of that computer was fabulous of its time, it was all discrete components.”

Brian’s thesis was ‘Studies in Laboratory Astrophysics’ and was based on early research which had tried to understand why the relative abundance of elements in the universe was as it is which involved understanding the nuclear collision processes that go on in the middle of stars. Brian looked at this, plus research of the 1940s which aimed to work out what the relative abundance of all the elements should be on the basis of nuclear cross-sections which had started being measured at that time. Brian explains: “There was a very seminal paper called ‘Burbidge, Burbidge, Hoyle and Narlikar’ which showed what the relative abundance should be based on the nuclear physics and the calculations.  There had been a lot of observations of similar quantities based on the atomic physics, because you only see the atomic properties in the outer layers of stars because that’s where it’s cold enough to see anything, you don’t see the nuclear processes, you only see the atomic processes, so similarly, people had inferred from measurements what the relative abundance of the elements should be. Most of them fitted, but one that didn’t was manganese, one of the iron group elements. My thesis examined why is it wrong, what is it that people have not done.”

From his own research, Brian discovered that people had ignored the fact that of all the elements in that group, manganese has a much bigger ratio of stable isotopes, which cause the spectral lines to be at slightly different wavelengths. He continues: “If you don’t take that into account in analysing the data, you end up with the wrong answer.  If you do it properly, which is what I did, you do end up with an answer which fitted the Burbidge, Burbidge, Hoyle and Narlikar equation.  I resolved the apparent paradox that the nuclear calculations and the atomic observations were inconsistent, they are consistent.  What is really nice is a lot of other people have done work after that with lots more sophisticated techniques than I had available in the mid-sixties and confirmed what I did was right.”

To complete his thesis, Brian did a lot of data processing and computer modelling, including building a computer model of the outer layers of the sun. He says: “It was a roll of punched tape, six inches in diameter.  I wrote it in ALGOL, I shun Fortran, I would not program in Fortran because it’s so clunky.  It was a very unsophisticated language.  I like the prettiness of ALGOL.” He did this work on the ICL KDF9 which was replaced in the third year of his thesis by a 1906 computer. Brian adds: “The 1906 didn’t have an ALGOL compiler, or at least it didn’t have one working when it arrived.” Brian was told to rewrite all of his progams in Fortran, however, he decided to debug the ALGOL compiler and make it work. He also debugged GEORGE 3, the operating system, to get that to work the ALGOL compiler. He continues: “I got deep into the entrails of this machine, not because I was particularly interested, although it was fun to work out how these things worked, but to get my programs to run so I could finish my thesis.”

In 1973, following three years as a postdoc student, Brian, now married with a child and one on the way, decided he needed to find a job. He applied to five public sector research establishments including the Royal Armament Research and Development Establishment (RARDE), the Royal Aircraft Establishment, GCHQ, the National Physical Laboratory and the Royal Radar Establishment (RRE). He was offered jobs by all but the National Physical Laboratory.

Brian chose the Royal Radar Establishment in Malvern, where he started as a Higher Scientific Officer (HSO), saying: “They didn’t actually offer me more money, but they offered me a much more exciting opportunity with regard to what I was going to be working on, which was basically laser systems.”

Brian was involved with two major programmes: night vision equipment and a laser system. He explains: “I was involved in developing a laser range finder system that could detect and determine the range of a tank without itself being detected. “

He explains that he did this by “using very clever signal processing and a deep knowledge of how the atmospheric physics worked.”  A patent was developed and prototypes were made but it never went into actual production, Brian explains: “It never actually happened because by the time we’d actually got it developed it was fairly obvious that land battle in Europe was pretty unlikely and so the need to have this force multiplying effect disappeared.”

 The Malvern facility was a tri-service organisation and home to around five and a half thousand people at its peak, Brian adds: “We were basically the systems integrators of modern sensor signal processing using modern electronics, modern computing during the seventies and eighties.  We did a lot of civilian work on microwaves, which then got exploited in mobile phones.  A lot of the early microwave stuff came out of Malvern.”

During the fourteen years Brian spent at the Malvern facility, he witnessed the downsizing of the defence industry which saw the disruption of Ferranti and Plessey with a shift away from developing equipment to buying ‘off the shelf’ equipment which he believes can lead to the loss of military advantage. Brian says: “Some of it is still there because we’ve kept some of it in our aviation capability.  The Royal Air Force managed to keep all the night vision, night flying capability. … I was party to that particular programme with a colleague from the Royal Aircraft Establishment. The air force got wind of the fact that the army were capable of doing some of these things at night with what was called thermal imaging and they said why haven’t we got that in our aircraft?”

Together with Jeff Fellows, who was the Superintendent of the division at the Royal Aircraft Establishment responsible for avionics, Brian spent eighteen months refiguring the kit to work for RAF. As a result, night flying at low level at high speed is now normal for the Royal Air Force.

Brian adds: “It was the days when you had this strategic research and development capability focussed on defence and doing the integration of technologies, systems, and sub-systems that came from all over, to provide a capability that made a difference.  Without that, the Falklands would not have been a successful operation.  We were all involved in the Falklands. I think there were 15,000 people in the defence establishments and those 15,000 all got behind getting the Falklands sorted. All of us did lots of various things, some of which is still not in the public domain, which made a difference to whether or not we were successful, because the Prime Minister having said go do, everyone said how?  Well, we just did, but that was a standing army of capability that isn’t there now.”

Brian progressed from HSO to Deputy Director and took on a more managerial role responsible for a division of around thirty-five people. He says of the experience: “I look back sometimes and think that’s probably the most fun job I had because it was the division in which I’d worked. It consisted of thirty-five very bright people and I knew all of them very well, I’d grown up over the previous six or seven years with them.  So that meant they knew that I knew their business, so I learnt how to let them get on with it but provide them with the context in which I knew they would succeed because I’d been there myself. … I learnt pretty quickly that you weren’t managing the people, you were managing was the context in which they were going to become successful. You helped them, mentored them, dealt with their stresses and strains and some of their idiosyncrasies and occasionally you had to be fairly tough about that, but that didn’t appear to worry me particularly.” Brian was provided with professional management training which helped, he adds: “The Ministry of Defence was very good at that time in making sure that its professional people were capable of being professional.”

In 1986, Brian spent a year at the Royal College of Defence Studies. He explains: “I spent a year being educated in how the world works geopolitically around defence, intelligence, and that was crucial to everything else I did and a lot of other things that I did outside defence following that.”

 At RCDS classes were asked to give talks about what they did, and Brian talked about Star Wars which he had been involved in as a member of the UK team participation in the Strategic Defence Initiative. At that time US President Ronald Reagan, planned to spend three and a half billion dollars on a missile defence system to combat the long-range missiles from the Soviet Union. 

At the end of 1987, against the background of the Chernobyl disaster and the start of the breakdown of the Soviet Union, Brian moved to GCHQ in Cheltenham as Chief Scientist and Director of Technology, the sixth person to hold the post after the end of 1945.

GCHQ’s role is to gather intelligence from communication systems both those that are open and others that are hidden and where people are trying to keep secrets.

The then Director, Sir Peter Marychurch, had set out a challenge that would see GCHQ change from what it had been doing since 1945 to something much more dynamic. Brian was selected for the role because he could introduce an external viewpoint as someone who was external but had been cleared to know what GCHQ did, “but not its deep entrails”. Brian says, “Sir Marychurch told me that my job was to get the technology and the systems sorted to support the change. No pressure then.”

Brian describes GCHQ at the time: “Even at the end of the 1980s GCHQ was still a very secret organisation, no one would talk about what it did, and yet those of us who came from outside knew there were telecommunications technologies, computing technologies that were going to arrive during the next decade that would completely transform the way in which communications of the world worked.”

The culture at GCHQ was completely inward-looking with a focus on the talented people within the organisation. Brian says that when you go into work inside GCHQ, you disappear into a black hole, he adds: “In fact a lot of people called it the Black Hole of Cheltenham. Part of my job was to bring in the knowledge of what was happening outside, expose them to what that might do and then start saying are we actually going to be able to handle what we need to deliver, which is intelligence for the country.  We weren’t there to break codes, we weren’t there to develop technologies, we were there to provide advanced intelligence about threats to our national wellbeing. …  It took a little while for people to get their heads around the fact that they didn’t control the way in which all of this stuff was developed outside in the outer world, that electronics and telecommunications in the computing world had grown now to a stage where it was so widespread that it was going to have a huge influence on everybody’s life everywhere in the planet.”  While at GCHQ Brian also highlighted the potential impact of mobile telephony and subsequently started the process of how GCHQ dealt with it.

Another challenge Brian faced was changing the way of working, improving the end to end systems of how everything worked between the different teams at GCHQ. With a responsibility for over two and half thousand people, Brian realised the scale of the issue when he pulled together his six direct reports who had previously never held meetings together.

On GCHQ’s ability to recruit talented people, Brian says: “The cachet of being at the sharp end of doing this type of work attracted some seriously good brains.  It had always attracted some of the best mathematicians because of the code breaking stuff, but we attracted some of the best electronics guys and I broke the mould by doing more external work and with contractors who I knew from my RSRE days had got some seriously good capability as a result a few pockets of industrial capability were developed, which have grown now to become part of their armoury.”

In 1991, at the age of forty-five, and after three and a half years at GCHQ, Brian’s secondment ended, and he left. With the privatisation of the various research centres and agencies, Brian found that the two or three roles that he could have taken up as a major-general, air vice marshal equivalent, had vanished. He turned to the private sector instead and became a partner at KPMG where he set up an information security practice.

He says: “I could see that unless the commercial world woke up to the fact, it wasn’t just going to be GCHQ that was going to be interested in what they were doing, everybody else was going to have the capability to find out what they were doing by electronic means, therefore you needed good information security.”

Unfortunately, the role at KPMG did not work out for Brian and he became Head of IT at the Wellcome Trust. He says: “The Wellcome Trust is a science-based organisation and the largest medical charity in the world and being a scientist that was quite attractive, and it needed a modernisation of its IT function.”

As well as updating the IT systems, Brian was also asked to look after the programme management of the IT development of Hinxton Hall, outside Cambridge University, where the Wellcome Trust was investing in the human genome mapping programme.

In 1988, Brian acted as a consultant for Clifford Chance, which at that time was the largest law firm in the world. Brian was initially contacted by them when a friend suggested he could help them with a security incident. Afterwards he was invited to run an information security review of the whole organisation in preparation for a merger with American and German law firms and he subsequently acted as the Global CIO managing the systems integration for the merger. His role saw him working with the IT teams across the organisation’s offices around the globe. In 2001, following 9/11 and the economic downturn, the company switched from operating at this global level to a regional level and Brian decided to leave.

Between 2003 and 2011 Brian was Professor of Information Systems at Cranfield University in Shrivenham, Wiltshire. He was part of the team that won the MOD contract for delivering university education at MOD.

Brian has been Chief Scientific Advisor for Department for Transport (2006-2011) and the Department of Business, Innovation and Skills (2009 and 2011).

In 2006, Brian had a call from Sir David King’s office, the government Chief Scientific Adviser, inviting Brian to apply for the role of Deputy Chief Scientific Adviser at the Department of Transport. Among many other projects while in office,

His role saw him dealing not only with IT but also with big infrastructure issues. He says: “I knew how the defence and intelligence world works because I’d been at the top table. I’d seen bits of the finance industry from the law firm and I’d been in bits of academia from various other places.  So, by the time I got to this position in the Department for Transport I was back at the government top table, but with a whole raft of different attributes. Dave King asked me in the interview, what would I bring to his party, and I said being a polymath which I know it’s not a particularly fashionable discipline.  He said, in this job it is, because that’s what you’ve got to be able to do, get on with everybody, understand where they’re all coming from and try and synthesise what it is they’re about.”

Brian identifies several achievements during his time in the departments including his work to improve the use of science and engineering evidence in policy making, coping with the Icelandic volcano eruption, creating a network of DCSAs, and putting engineering evidence on a firm footing.

In 2011, Brian became the Professor of Engineering Policy at University College London.  Brian says: “At the moment we are attracting some of the brightest and the best and some of the young people giving me hope for the future because they actually get it.  They’re not on ego trips, they see all the big issues that are out there: climate change, demography, resource scarcity and the impact of all of that on ecosystems and how we live.”

On the subject of HS2 which at the time of Brian’s interview under review as to whether it should go ahead, Brian says his decision would be “don’t build it.” He goes on to explain: “Not as it is.  You increase the capacity of the railway system between the north and the south, but I think what you should do is put a big pause on it being called HS2, because it’s not high speed. Instead, the Government should say now we’re going to develop the capacity for improving the rail communications between Birmingham, Manchester, Leeds and London and we want to see what the demand is to determine where those first links should be put.”

When asked what role IT plays in this, Brian says: “It’s a digital railway, just like the Queen Elizabeth line is a digital railway.” On the issues that have delayed the Queen Elizabeth line, Brian points to a lack of integration in the way that the programme has been considered from the outset, he says: “It’s not been governed from end to end as a digital railway which happens to go through tunnels.  It’s been governed as three massive contracts; civil engineering, railway, and digital, all separated from each other.  The big issue is that there has been no systems thinking, no system or systems engineering around the whole thing end to end.”

Brian says on the subject of perceived public sector IT failures: “The media go after public sector failures of any sort, IT or otherwise.  They never give public domain visibility to those things which are successful because it’s not good media, it doesn’t sell papers or news programmes.  So, I think I’d temper the allegation that it’s not very good.”

He also points to the fact that public sector projects have financial scrutiny through the Public Accounts Committee which large, commercial companies rarely face in the same way. He adds: “So because there’s a record you can see how well it’s been done, or it hasn’t been done.  If you go and talk to the private sector there isn’t a record of how well it’s been done or not done.  So, I don’t think it’s anything like as bad.

“I also think we’ve managed to lose an enormous amount of the expertise on the client side of the procurement activity. People are talking about the public sector becoming more of an intelligent buyer, an intelligent client in the procurement of IT, and a lot of other things, because at the moment they’re not.  They’re informed.  There’s a phrase that’s used, informed, intelligent and expert.  It’s very expensive to be an expert buyer but occasionally you have to do that for example, if you were buying very special drugs you probably have to be an expert buyer, you have to know exactly what it is you’re asking for, but if you collapse everything down to informed, then you get what you pay for. When IT is such a critical factor in the development of and operation and delivery of public services, then you would argue that you should be an intelligent customer.”

Brian says that he has some concerns about the development of artificial intelligence because it is a big area of unknown, he adds: “but we’ve faced big areas of unknown in the past.  When we build nuclear bombs and nuclear power stations we didn’t know how the nuclear physics worked. …  I’m not over concerned as long as the A star team is addressing the issue. … If in any way the governance of the outcome of all of this is devalued or made mediocre, then I’d be more worried.  … We’ve gone through a number of technologies which have had that sort of impact on us and if you look at the history over the whole of the Industrial Revolution, all of those things of a different nature have caused anxiety in the fabric of society. … It’s part of evolution, I guess.”

On the subject of Y2K, Brian says: “Lots went on under the bonnet that people never talked about as to what they corrected.  One of the main outcomes was that we found out what our system configuration is, made sure all operating systems are up to date, made sure all the version controls are right, made sure all the connections are right. Who’s accessing your machine is right, because those are the people who import bugs that you wouldn’t otherwise know about.  So, there was a huge amount of systematic scrutiny, what nowadays is more routinely done and is called basic housekeeping.  A lot of people, particularly in the finance sector, engineering and defence got their act together in advance and discovered all sorts of things that would go wrong, not necessarily millennium bug things that would go wrong, and as a result the whole quality of IT came up.  Now is that an audit that anyone would talk about in millennium bug terms?  No.”

Brian is a Fellow of the Royal Academy of Engineering, the Institute of Civil Engineering, BCS, The Chartered Institute for IT, the Institution of Engineering and Technology, the Institute of Physics, the Royal Society for the Encouragement of Arts, Manufactures and Commerce. He is also a member of the Royal College of Defence Studies.

Interviewed By: Richard Sharpe on the 28th January 2020 at the Worshipful Company of Information Technologists’ Hall

Transcribed By: Susan Nicholls

Abstracted By: Lynda Feeley