The Innovation Highway project will utilise AI and machine-learning to optimise the full innovation value chain. The platform developed will help facilitate collaboration amongst networks, and other sectors such as water companies so they can innovate together. AI-empowered algorithms will simplify the identification, mapping, assessment and selection of problems and ideas, reducing manual processing time and enhancing effective decision making; this will support identifying and prioritising projects that will deliver the highest benefits. The platform will also help networks automate the development of cost benefit analysis.
Benefits
1. Reduction in time needed to process ideas
This benefit is based upon the time saved engaging with innovators unnecessarily and sifting through ideas prior to selection of what should and shouldn’t be approved and developed into a project. This would be informed by insights from previous projects, identification of similar historical ideas, and potential solutions that already exist.
The calculation of this benefit is based on the average number of ideas that have been rejected over 2021/2022. The project will reduce the time taken by an innovation engineer to review and reject these ideas by 60%. This is a recurring annual benefit.
2. Reduction in number of ideas unnecessarily progressed through governance
The project will create a 60% efficiency saving from investing resource in taking ideas through governance that then get rejected. The annual saving is based on the average number of ideas developed through Gate A and Gate B (these are UK Power Networks’ internal governance stage gates) that did not commence into full projects over 2019-2022.
3. Reduction in cost to develop project CBAs
The project will create a 60% efficiency saving via the AI tech flex that would look to automate the development of CBAs, reducing the need for human involvement. The cost saving is based on the average time spent creating CBAs over 2019-2022, assuming a continuous annual cost saving after deployment.
4. Uneconomic Project Spend Reduction
Over ED1, based on E6 reports, there was a total of £4.95 million spent on projects that returned no benefits. The project will reduce the spend on projects that return no benefits by 40% by focusing on projects that are more likely to deliver value through better problem identification and ideation. This will result in more projects having benefits.
For UK Power Networks, this could result in a £0.83m saving over RIIO-ED2
Base cost (PV) £3.76m
Method cost (PV) £2.93m
For UK Power Networks, this could result in a £1.63m saving over RIIO-ED3
Base cost (PV) £4.00m
Method cost (PV) £2.37m
Similar benefits are anticipated for the other participating networks, where the value will be based on their relevant size compared to UK Power Networks.
These benefits would apply across the entire innovation portfolio whether funded via the Network Innovation Allowance, the Strategic Innovation Fund or business funded innovation. Further refinement and communication of benefits will be undertaken during the Feasibility stage (Stage 1) of the project.
Learnings
Outcomes
At the beginning of the project the method was assessed as being at a TRL of 4. After the development of the method throughout the project, it was possible to use the platform to facilitate collaboration in genuine working environments such as industry conferences. On this basis, the method can be considered to have advanced to TRL 6 -7. This has been achieved by creating a solid foundation based on technical advancement and thorough understanding of user requirements, and how the product would be used. Some of the figures that demonstrate this advancement are:
Development of 33 business use cases and 66 gen AI use cases
Creation of 319 user stories and 47 epics
Capture of 71 functional requirements and 14 non-functional requirements
In addition, the method has been enhanced by the design of sprints and epics for future development of such a platform, low-fidelity designs for new modules that could be created to further enhance the system as a commercial product, definition of the system and data architectures. The AI solution design was also created, evolving the technology innovation aspect of the project. Understanding of how the product would look and feel, and how users would need to engage was also advanced. This was achieved through wireframing, behavioural science audit, user testing and prototype creation.
These outputs from the project demonstrate the increased sophistication of the method which translates into the higher TRL. But, beyond this, the project has advanced the understanding of how advanced technologies can be used to facilitate collaboration not just within the industry, but to identify cross-industry challenges where, previously, it may not have been possible to identify where different business types had similar problem that they could work together on to solve. It has led to a tangible roadmap to the delivery of such systems and an understanding of what people need from them. This provides a firm basis to advance learnings in future project, which could take the concept further.
Lessons Learnt
Further module development and testing would be required in order to develop a commercial-grade MVP. This would very much be around user trials in which key stakeholders use the solution and provide feedback on how well it suits their needs, what gaps exist, and what improvements could be made. This form of trialling would ensure alignment user requirements.
Problems with the trialled Methods: The methods deployed through the first two phases of the project proved effective in achieving the objectives that were defined, and delivered against the success criteria outlined. There were therefore no issues with the trialled methods, and so could be replicated successfully.
Effectiveness of any Research, Development or Demonstration: Research, development and demonstration were all key elements within the project. Development and demonstration were crucial to the production of the prototype, and the resulting feedback elicited from stakeholders upon demonstration were fundamental in achieving the objectives and success criteria, in particular ensuring alignment with user needs and development of the future roadmap.
