Defenders of knowledge

For the UK government’s Defence Science and Technology Laboratory, research and information-gathering facilities are critical to developing defence innovations. Steve Thornton explains the role of knowledge agents in the creation and preservation of a knowledge-sharing culture within the organisation.

At the Defence Science and Technology Laboratory (Dstl), a ‘knowledge agent’ is a trained information generalist: an intermediary whose function is to provide expert advice and support for research and analytical projects. The success of knowledge agents over the past three years has taken all those involved by surprise.

Background and origins

The background to the knowledge-agent concept is largely based on my personal experiences and career in defence research libraries. As a research librarian, my career has spanned over 30 years working for organisations that include the Admiralty Research Laboratory (ARL), Royal Aircraft Establishment (RAE), Defence Research Agency (DRA), Defence Evaluation and Research Agency (DERA), and, currently, the Dstl. Over the years I have had the opportunity to provide services to a wide range of scientists and technologists. Often among the top five or ten in their respective fields, a large number of these professionals have played a part in developing inventions that affect our lives today, including the jet engine, radar, automatic landing, the LCD display and carbon fibre, to name a few.

Early on in my career, I discovered that sitting in a library wasn’t much fun. In fact, it was extremely boring and I found myself walking the corridors, visiting customers in their offices and chatting to them about their current projects. Through these informal chats I gained invaluable insight into information-gathering habits and the information needs of scientists. In addition to this, I became involved in these projects, not simply as a passive supplier of books, journals and reports, but as an active – if unofficial – member of the team.

I learned that scientists were only too willing to talk to a friendly librarian, even if he knew nothing about science. I also realised that it was often possible to identify information needs in areas where I could assist and that many of these needs would never have been vocalised if I hadn’t been present. This awareness enabled me to keep an eye open for anything new that might come through my hands and, in turn, made my role more satisfying.

As I moved from ARL to RAE Farnborough and on to RAE Bedford, I was able to pursue this radical approach. However, more of my time now seemed to be spent on management as opposed to librarianship. In 1992, the government embarked on a massive re-organisation of defence research beginning with the merging of the major establishments into DRA and then into DERA in 1996. Although initially ignored as overheads, the libraries were finally drawn into a single management structure in 1994. To begin with, each of the existing site chief librarians were treated as equals, but it soon became obvious that many of the libraries offered different qualities of service to their users. I was tasked with the role of ‘quality’ and later ‘improvements manager’, with a remit to improve the level of service across the board.

It became obvious that many of the senior staff in the libraries saw their function as keepers and builders of collections, rather than providers of services focused on user needs. Staff cuts had led to unsuitable leaders being put in charge of services, and retrenchment had merely set them more firmly into that mode. This did not make them bad librarians, merely unsuitable for the tasks that were now required of them. With this lack of dynamic staff, improving services became a nightmare.

From data miners to DERA knowledge agents

In April 1998, our new head of sector indicated his support for expanding outside the boundaries of conventional library services. In addition, talks with other senior staff had made it apparent that project managers felt that there was a vast wealth of untapped material that could be of value to them. Staff also felt that they could not manage their own project material easily and that they lacked expertise in both information and knowledge management.

There were very few among the existing library staff who could actively support these needs. My original idea had been to simply employ additional assistant librarians, but I became convinced that this would only lead to an increased number of staff spending time on routine library tasks that had limited value, and hence fail to reach out to the potential community of users that required a broader range of information services.

To counter this shortfall, in May 1998 I proposed the recruitment of a team of ‘data miners’, who would have a separate identity from their conventional counterparts in the libraries. Each data miner was to be a resource of the regional manager rather than the local library, have a separate, specific job title and have a seat in the local library, but no more. The duties and skills necessary for the role were:

Duties:

To support specific projects when they were assigned to the tasks by their regional manager;
To be aware of the full nature of the project and attend all relevant meetings;
To gather data from a wide variety of sources in support of the project;
To organise the project’s documentation when required;
To become an integral part of the project team;
To act as a conduit into the project for our other services;
To have a wide-ranging knowledge of the facilities and staff available in the department to turn to for expert advice and skills;
To be available for local duties only if and when project duties permitted.

Skills and talents:

A good general education (valued above specialist knowledge);
Training in information retrieval and/or document management;
A lively and enquiring mind;
The desire to be part of a team;
A dynamic and self-motivating nature;
A couple of years’ experience;
In-depth knowledge of the internet.

These proposals were accepted, although Chrissie McCracken, the departmental head, changed the job title from data miner to knowledge agent, a term the Dstl has used ever since.

Recruitment went ahead and the posts were advertised in the professional library and information press, as well as in the in-house job-vacancy newsletter. We didn’t expect any response from the in-house newsletter, as we knew all of the librarians who might be eligible and did not feel that any of them would apply. However, we did receive an enquiry from a chemist whose project-based role had been similar to our specification. Knowing that detailed knowledge of chemistry literature was a major shortfall in our department’s skill set, we advised him to apply. Although we didn’t realise it at the time, this was a critical turning point.

Our intention was to recruit four members of staff but there were six suitable candidates, including our chemist, all of whom were employed. Recruiting these staff was a brave decision, as the whole initiative was a financial risk. The knowledge agents would have to earn their keep from project budgets and it was anticipated that it would take some time before these projects were up and running. They would also require expensive training.

After six months, the knowledge agents were ready to begin. We gave them a broad remit of responsibilities: to find people in the departments who needed their professional help and to provide assistance. A round of presentations to departments, sections, project teams, research into what potential customers would be interested in and various training initiatives eventually saw work trickling in.

Take-up was initially slow and it proved a difficult market to penetrate. But after one or two successful projects, the knowledge agents were starting to be noticed. One success story was their involvement in a short-staffed, urgent project for the Cayman Islands government, which the knowledge agents handled expertly. The project came in on time and under budget, and the agents were mentioned favourably.

Just as the knowledge-agent concept was starting to take-off, it was announced that DERA would be partially privatised and that further recruitment was prohibited. At this time, the knowledge agents were immersed in a major internal project that made huge demands on their time and resources. The agents were running themselves ragged trying to meet targets and promotion was no longer a possibility.

Privatisation was planned for July 2001. Three-quarters of DERA was to become a privatised company, QinetiQ, and the remainder would stay within the MOD, ultimately becoming the Defence Science and Technology Laboratory. Chrissie McCracken was invited to become the head of Dstl’s new Knowledge Services department and I was invited to become technical manager.

Dstl gets its own knowledge agents

Our Dstl department consisted of some managers divided between the Farnborough, Bedford and Glasgow sites, a small but highly effective specialist research library at Porton Down, and two co-located teams that had formed the Defence Research Information Centre in Glasgow. Although about a quarter of Dstl’s proposed staff were on the Porton site, there were only four other Dstl staff in Scotland aside from the Glasgow team. Approximately 2,500 of our 3,200 staff scattered on 12 other sites therefore had no library and information services at hand, and no knowledge agents. The old libraries and the entire knowledge-agent team had been earmarked for QinetiQ, so Dstl Knowledge Services had to start from scratch.

I was tasked with helping to define and refine the services our new user community required and with working on how to meet their needs in a way that Dstl could afford and would be willing to pay for. Senior management at Dstl approached the project with a vastly different mindset from that of DERA. DERA was a very commercially minded organisation, aimed specifically at reducing costs and maximising profit. While this was laudable, this approach could be detrimental to information services and the effective use of knowledge skills within the organisation. It soon became policy that everything over and above a very limited service was charged for. This led to a steady change in individuals’ perceptions: money spent on knowledge-agent services meant there was less money in the pot that could be spent within each department’s profit centre for staff deployment or new equipment. In some areas this led to inappropriately trained staff being used to gather information inefficiently, and a reduction in relevant articles and other information being acquired. Some departments actively opposed this, but they tended to be in the minority, generally because they lacked the staff time to waste in such tasks.

From the beginning, Dstl’s management made it clear that excellence and cost-effective, timely support for the MOD were of higher immediate priority than quarterly bottom lines. Indeed, if an MOD customer asked for work to be undertaken that had been done before, or could be done more cheaply, the Knowledge Services department would say so, suggesting other areas where money could be spent more effectively.

In addition to this sea-change approach, management also made it quite clear that the information professionals under the knowledge-services umbrella were perceived as the experts in the field, and that our opinions, suggestions and recommendations carried significant weight. In many cases, work in this field could often lead to relevant information and knowledge-related programmes. This attitude towards knowledge work was a refreshing change.

Therefore, the re-introduction of knowledge agents was a priority. Although no new staff could be taken on until 1 July 2001 (when Dstl was officially created), the Knowledge Services department had already launched a recruitment campaign, with adverts in the professional information press and in New Scientist. Previous experience with our original chemist, as well as subsequent DERA knowledge-agent recruits, had reversed our attitude to employing scientists. We were now looking for was either librarianship/information training or evidence of significant interest in this area. The response was staggering. From a total of 158 applications, over 90 per cent were from scientists. We were again only intending to recruit four knowledge agents and were prepared only to employ the best. If we didn’t get suitable candidates, we would start from scratch. Luckily the overall quality of the candidates was impressive and we ended up with six recruits. Only one had librarianship training, but the others were very well educated, demonstrated a comprehensive awareness of the value of information and were aware of the sources available, making training easier.

The remit was the same but our new knowledge agents were very much on their own. We wanted employees who could hit the ground running, and we were not disappointed. Our last starter described his initial experience as “not so much a learning curve, more a learning cliff, with overhangs”. Working closely with a team of information scientists in Glasgow, the knowledge agents began to pick up small tasks, building up their own lists of contacts. Some of our clients had already experienced DERA’s knowledge agents and so take up was quicker than we expected: users were aware of the importance of ‘buying in’ expert support rather than attempting a DIY job and wasting valuable time.

I was able to suggest the use of knowledge agents to a project manager who was proposing a pan-Dstl capability analysis to the board, but was finding that there was lack of trained staff to conduct the analysis. The board approved his proposal and supported the idea of employing knowledge agents to support his endeavour. Over the next three months, the majority of the knowledge-agent team interviewed the top-five levels of management within Dstl, building up knowledge of the organisation that would normally have taken years to accumulate. Demand for these services has grown exponentially and we currently employ 18 knowledge agents, with plenty of projects in the pipeline.

Assessing the agents

In autumn last year, as part of Dstl’s Technical Excellence programme, the knowledge-agent team was assessed by external experts. Although they identified some weaknesses (encountered by all new teams) and criticised the ‘go and do’ management style, the concept and the knowledge agents themselves came in for almost undiluted praise. “Overall, the knowledge-agent concept is clearly a brilliant innovation and a potential model for other areas of the public sector. We were very impressed with the quality of the individuals it has attracted and firmly believe that the knowledge agents could have a major positive impact on Dstl and more widely across the MOD.” This was just one of the positive comments offered by the experts. Following the assessment, the knowledge agents were presented with the ‘Best information/knowledge team in the public sector’ award at the International Information Industry Awards in 2003.

The fundamental principle of the team remains the same – to provide proactive and intelligent information and intermediary support to projects – but has grown from the original concept. The knowledge agents have developed skills and techniques far beyond expectations and are now developing an underlying theory and philosophy of ‘knowledge agency’.

The knowledge agents at the Dstl now provide everything from a basic literature search to an organisation-wide capability analysis; from identifying an obscure reference to an in-depth review of literature; from facilitating meetings to applying data visualisation to technology tracking. They are beginning to become integrated into other areas within the department, including mentoring members of the RCS in Dstl and gathering information for a joint MOD/US DoD Committee. The knowledge agent is not simply a scientific discovery; it has become a variable phenomenon without limitations.

Steve Thornton, departmental programmes co-ordinator, Dstl Knowledge Services, sathornton1@mail.dstl.gov.uk