## Engaging with industry - An Occasional Blog by Paul Glendinning, Scientific Director

ICMS is currently hosting a meeting of the UK Mathematical Sciences Knowledge Exchange community to talk about strategies to make sure that there are efficient pathways for the exchange and development of ideas between academic mathematical scientists and the industrial and public sector users of mathematics. Philip Bond is here to listen as part of the Government’s Review of Knowledge Exchange in the Mathematical Sciences which he chairs. Earlier in the summer the Review made a call for evidence and ICMS submitted the response below. For me the important features are reach to the whole mathematical science community and embedding industrial activity as standard practice in all departments. As part of the culture change that this implies, we also need straightforward (e.g. NOT involving industrial placements which are hard to scale with class size) undergraduate courses explaining the R&D process together with examples of the use of mathematics and how it came to be part of a product. Only by making the application of mathematical sciences in society (both the how and the what) one of the ways all mathematical scientists are trained to see their subject will we really start changing attitudes.

*Response to Review of
Knowledge Exchange in the Mathematical Sciences request for evidence.*

International Centre for Mathematical Sciences (ICMS), Edinburgh

Professor Paul Glendinning (Scientific Director), Professor Gabriel Lord (Chair of ICMS Management Committee) and Dawn Wasley (ICMS Knowledge Exchange Officer)

**1. What are the strengths and
weaknesses of the current support mechanisms in place in the UK to assist in
knowledge exchange in the mathematical sciences?**

Strengths:

- There is a growing academic mathematical science community interested in industrial problems (see attendance at the recent Royal Society meeting, Mathematics for the Modern Economy) and broader recognition that industrial problems can lead to genuinely new mathematics.
- There is infrastructure support through the Newton Institute and Turing Gateway to Mathematics (strong contacts with industry, themed meetings, connecting with research programmes) and ICMS (meetings to establish good practice, nurture of developing communities, connecting to research workshops, small group research funding).
- Smith Institute, KTN and InnovateUK provide strong support for those who are part of their network.
- Other institutes such as Alan Turing Institute are useful to those directly involved. Internal community support (Modelling Camps, Study Groups with Industry) also help.
- University support (appointments in industrial mathematics, centrally led cross-disciplinary activity) is growing.

Weaknesses:

- There is over-reliance on a small number of active individuals to drive the agenda. This means that connections with industry are very non-uniform across the country. This small number can be seen as a clique – there is a risk others, active in engagement do not actively drive the agenda. ICMS can help by bringing a broader section of the community together.
- Targeted academic appointments are hard for smaller departments.
- It can be hard for individual academics to know where to start (do they have the right expertise, how do they find contacts?) and no well recognised route to industrial collaboration.
- There are no uniform mechanisms for the mathematical sciences community to provide collective responses to new initiatives.
- It is often difficult to track results of industrial collaborations and influence (follow up is often not a high priority for researchers and there seems to be no central evaluation methodology).
- There is a perception that industrial mathematics is as not as valid or important as traditional mathematics and publications are not valued in the same way. This can lead to issues around career advancement and engagement with ECRs.
- Emphasis on more immediate impact means longer term mathematical analysis is not prioritized or funded. This means more engineering type solutions are promoted.

**2. How is
mathematical knowledge exchange currently incentivised and are there any more
effective ways to incentivise knowledge exchange?**

REF has provided a big incentive for universities, though the narrow definition of impact makes it hard to recognise the full external impact of mathematical sciences. Some academics now have ‘industrial’ in their job description and hence in their promotion criteria. EPSRC ‘pathways to impact’ emphasises the importance of KE (where appropriate to the proposal), hence effective KE activity helps grant successes.

In the medium term, the professionalisation of academic mathematical science knowledge exchange (people with explicit KE responsibilities in their job titles or duties) in universities and the mathematics infrastructure will lead to more interactions and those interactions will be more effective. Removing the perceived barriers to participation (communication and information) will help mathematical scientists to develop in these areas as part of their career choices. From the point of view of industry, tax incentives would help companies to invest in longer term research projects in the mathematical sciences (cf. USA) and help invest in the students who undertake some of the work and will be employed in those industries in the future. We need to challenge (communication and information again) the perception that mathematics is NOT at the heart of industry (Deloittes report, fact that mathematicians in industry are often labelled as engineers or computer scientists). Provide more accounts of how mathematical sciences has made a difference to industry and the public sector.

**3. Are there
any challenges to knowledge exchange which are unique to the mathematical
sciences? What are they? Please also identify opportunities to learn from
approaches to knowledge exchange in other disciplines.**

External perception is a problem (see 2 above). Innovative mathematical science approaches are more risky than engineering or computer science approaches. The latter tend to see it as a success to do something a bit faster or a bit bigger (see 4a below), but a new mathematical approach to a problem may not work! However, when it does it can make a big difference to the methodology used by economists, engineers and computer scientists.

The mathematics community does not have the resources of (say) the Institute of Physics or Royal Society of Engineering to provide regular publicity to the media about successes. It is hard to see how this can be changed in the short term, but the communication of the successful role of mathematics ‘under the bonnet’ is important, and the fact that mathematical scientists are increasingly effective at working with industrial partners also needs more visibility at all levels.

We do not as a community have a pool of PDRAs on tap and able to take up the often short timescale positions as traditionally there have been few opportunities. This means there is often a lack of available skills to take, for example, the IAA offered through the universities. Such short timescale positions all makes it hard for UK PDRAs to compete on the international market when it comes to permanent positions.

**4. What are the key opportunities
for the UK economy, society, people and knowledge associated with effective
knowledge exchange in the mathematical sciences? What do you think the missed
opportunities could be? When answering the above, please consider:**

*What are the current
opportunities?*

- The Review
of Knowledge Exchange in the Mathematical Sciences can create more coherent pathways
to more effective engagement
*from both sides*(academic and non-academic). - The community need to realize that small efficiency gains (e.g. numerical analysis methods) can lead to significant financial gains for companies.
- The Government currently recognises the importance of research agenda – mathematics is both need by this agenda and the mathematical sciences community needs to be part of that agenda. We need better ways of capturing this.
- The centrality of data science, including machine learning, to many business and health initiatives means that mathematics should be equally central (see e.g. ATI for both positives and negatives!).

*What are the opportunities looking to the future?*

- The creation of recognised pathways for industrial collaboration available to all (both sides, without specialist KE knowledge).
- Embedding industrial mathematics into the undergraduate curriculum so that all mathematicians have the opportunity to appreciate the possibilities of mathematics.
- Providing Modelling Camps for all postgraduate students (providing hands on experience, greater employability, and more opportunities to interact outside academia).
- Creation of a central voice for mathematical sciences to make recommendations for the role of mathematics in new initiatives and the processes whereby mathematical sciences interact with other disciplines.
- A significant increase in mathematical research funding across all areas to ensure the level of expertise in the area needed by innovation led research.
- A significant increase in industrial mathematics funding through greater understanding of the impact mathematics can make.

**
5. Do you have any other comments
you wish to feed into the review?**

No.

**Return to ICMS Blog postings list.**