Smart Cities

Innovate UK SBRI: Improving Business Operations Through Machine Learning

Improving Business Operations Through Machine Learning featured image

Innovate UK is to invest up to £250,000 in new ideas for ways in which machine learning can improve the efficiency and effectiveness of its operations. Machine learning is a type of artificial intelligence. It allows computers to become more accurate in predicting outcomes without explicit programming, using algorithms that iteratively learn from data.

Analytics Engines have just won a contract for phase 1 of this SBRI to build a new software tool, Cobalt Grant Manager. It will utilise machine learning to help understand economic exploitation data to better inform investment decisions, improve the accuracy of expenditure claims forecasting in order to schedule projects more effectively.

Innovate UK will play an integral role in the forthcoming UK National Innovation Plan having responsibility to de-risk, enable and support innovation within the UK.  To meet this responsibility, a key objective for Innovate UK is to maximise the benefit of innovation for the UK economy into the next decade. This requires Innovate UK to understand the economic impact of their funding and to control their budget well so as to avoid missed opportunities.

Having invested around £1.8 billion in innovation since 2007, Innovate UK has a need to understand the economic impact of their investments to ensure the quality of future investments. Ten years’ worth of investments means that data is available to analyse not only short-term effects of funding, but also mid- and longer-term impacts, such as company growth indicators. Application of machine learning to this data will provide insights into economic exploitation and help inform future decisions.

Innovate UK is tasked with not over-spending, but also not under-spending by more than 1%, against allocated core budget, which leaves little margin for error in the spend targets and budget control.  However, grant recipient expenditure claim forecasts are not typically accurate to within the 1% variance, but rather around 10%. A 10% underspend across numerous projects results in substantial missed opportunities.  For example a 10% error across the core budget of £550 million would be £55.0 million, an amount that could have funded additional innovation projects. With no expected improvement in current forecasts, this problem needs an innovative solution to help tackle it.

Using our data platform, Analytics Engines XDP, we will bring together data and systems from across Innovate UK, third parties and open data, creating a single view of all the information available.  Then using inbuilt advanced analytics capabilities, we will apply machine learning and predictive models to aid in tackling these problems.
The purpose of the project is to develop a tool that will automatically provide Innovate UK with real-time monitoring information about their investments and costs.  We will also build predictive models based on recipient forecasts, historical data, business/organisation data and other data to predict actual costs and economic impact of investments. This will allow Innovate UK to make more  informed investment decisions and to take action to minimise underspend.
The tool will be based on proven machine learning techniques using large volumes of data to identify patterns and categorise events from disparate data sets, becoming more accurate and efficient as the models are trained and the data available increases.

Back to Case Studies

Let’s talk about how Analytics Engines can help improve your business.

Learn a new approach to big data analytics. Get in touch to discuss how we can help.

Contact Us