Knowledge Exchange (KE) involves university staff interacting with a range of nonacademic audiences to either co-produce information or exchange knowledge between them. This often takes the form of consultancy, research, CPD and training with businesses and organisations, but also includes licensing, company creation and enterprise, as well as informing policy and practice, and public engagement.
Knowledge exchange activities encourage the sharing of ideas, data, experience, and expertise, which is mutually beneficial to all parties involved. Engaging with external organisations often generates new ideas, creating opportunities to explore new avenues for research.
KE is one of our key activities as a member of staff at a university. It applies mostly, but not only to academic staff. It is a key indicator of our success and is assessed in our ADR (annual development review) and in promotion cases. It is often a desirable, sometimes essential criterion in a variety of job descriptions.
Funding for KE activity can come from a variety of sources, including innovation vouchers, Knowledge Transfer Partnerships (KTPs), Impact Acceleration Accounts (IAA) and consultancy and commercial funding. Visit the RKES SharePoint portal (log in required) for information on how to source different types of funding for research and knowledge exchange activities.
Below are some case studies:
Thales
Can humans trust and work with Artificial Intelligence?
Industry: Transportation/Defence.
Benefits: Better understanding how human operators interact with AI agents in complex decision-making situations.
Type of engagement: Industrial PhD.
Challenge: Software AI agents are embedded within safety-critical environments that involve complex decisions, where a single mistake could mean the difference between life and death. Thales wants to better understand how people trust and interact effectively with such AI-based agents.
Approach: A joint industrial PhD applies world-leading research from CIS on effective experimental design to explore questions around interactive information retrieval and human computer interaction to key questions of interest for Thales.
Outcome: This research will provide new approaches to assessing multi-dimensional data, by providing intelligent agents that a human can trust and effectively collaborate with. Such agents can support collaborative information interaction and facilitate complex task completion, which has application to domains such as air traffic control, sonar operation, complex surgical tasks, etc.
National Physical Laboratory
Making it easier to develop trustworthy software
Industry: Measurement/Engineering.
Benefits: Appreciation of formal methods and program verification leads to increased trust in software.
Type of engagement: Funded research and funded PhDs.
Challenge: Scientific computing programs can be wrong in subtle ways, with the intent of the software only made clear in the comments section, rather than the code itself. NPL would like to better guarantees that the software actually performs the required tasks.
Approach: CIS developed a state-of-the-art software verification system that can read both code and formal comments, and assess to what extent the code is consistent with any constraints implied by the comments, for example relating to units of measure. The system can also be directed to automatically generate code when the intent in the comments uniquely describes its behaviour.
Outcome: The verification system is available for NPL to use in their software development, and NPL employees also have an increased awareness of issues of software correctness. Planning is underway to seek further external funding to continue the collaboration.
Welbot
Smart technologies to support health in the workplace
Industry: Healthcare/Software.
Benefits: A clearer understanding of how users interact with a wellness application leads to more effective ways to deliver healthy lifestyle advice.
Type of engagement: Basic and Advanced Innovation Vouchers.
Challenge: Welbot's primary product is a workplace-based, wellness desktop application that sends wellbeing notifications throughout the day. The challenge was in timing the notifications to interrupt without disrupting: to improve productivity not diminish performance.
Approach: Welbot selected CIS based on their experience in analytics and AI. Supported by an Innovation Voucher, the work focused on understanding the nature of users’ interactions and enabled data models and logic to be developed, and motivational tools to be incorporated into the application.
Outcome: Welbot sought additional advice in exercise physiology, behavioural analysis and change. Experts from Psychological Sciences and Health, as well as CIS, are now focusing on content creation, and introducing personalisation, and context awareness to the product.