Guest article from Dr Joanna Jordan and Dr Shirley Coleman
Maths and numeracy are receiving increasing interest. Given the vast applicability of these skills, there is curiosity and enthusiasm to grasp the opportunities. What are these opportunities? Quantifying, describing, and analysing data can reveal surprising insights and benefits which we aim to give a flavour of here.
I am a people person, and firmly believe that maths can change the world for the better, which is why I work in mathematical sciences knowledge exchange (KE)! My PhD was in noise and vibration problems, sponsored by an engineering company, which gave me my first taste of KE in action. I also really benefited from participating in European Study Groups – week-long problem scoping/solving workshops, like hackathons – with Industry.
Maths is a language
Fast forward 15 years and working at the interface between mathematical sciences and industry has given me a privileged view of where maths can add significant value. Many companies have challenges, which on first sight may not sound like mathematical ones; often the first task might be to turn a qualitative description such as “how can I make this process more efficient?” into a quantitative question.
As maths is a language, there is a high degree of transferability, and ideas and models developed for one process often find application in other, seemingly unrelated processes. Philips came to a Study Group I organised in Limerick in 2012, looking to improve their design of filter coffee machines. The approach my colleagues took was to adapt equations that described groundwater flow and the workings of lithium ion batteries!
Economic and societal impact
The 2012 Deloitte report estimated the contribution of mathematical science research to the UK economy in 2010 to be 2.8 million jobs (around 10% of all jobs in the UK) and £208 billion in terms of Gross Value Added (GVA) contribution (around 16 per cent of total UK GVA). I recently wrote about the impact of investment in mathematical sciences research which included using network analysis to model the spread of disease, developing mathematical models to personalise COVID-19 treatment, and leveraging social media data to help prevent crime.
Surprising applications of maths
The range of problems where maths can add real value never ceases to amaze me. This year’s Study Group in Edinburgh included forecasting tourism demand for TravelTech in Scotland, modelling flow in home dialysis machines for SME QuantaDT, and better understanding the cooking of crisps for PepsiCo. All these applications showed measurable benefits for the companies involved.
My focus is on statistics and data science applications and I work with industrial partners from a range of different sectors, often via knowledge transfer partnerships (KTPs). These have included:
Analysing fuel consumption data from passenger ferries, offshore vessels and pilot ships, to improve the services offered by a marine company, increase sales, exports and profits and create new jobs.
Improved accounting of gas in the national transmission system provided by National Grid. Combining decision trees with statistical models a source of unaccounted for gas was identified, resulting in £14 million being returned to the community.
Analysing millions of transaction records from social housing leading to new insight on how to help people manage their rental payments.
Maths can be combined with other subjects. For example, maths enthusiasts may have particular personality traits. How do these compare with the personalities of people working in different areas of a company, often collecting operational data as part of their jobs? Are there better or worse ways for maths enthusiasts to discuss data with business personnel?
A picture is worth a thousand words
Some forms of presentation appeal more than others. For example, the Pareto chart has been described as the best management tool ever invented. A pump manufacturer made a valuable finding when they plotted a Pareto chart of enquiry numbers of a range of different pumps and showed that most enquiries were for pumps held in stock. Only the more bespoke items had a lead time of up to six weeks. Previously, reception staff told enquirers that there would be a 6 week lead time for all sales, just to be on the safe side, and yet now they know that 80% of enquiries can be satisfied in a few days.
Maths projects aren’t confined to summaries and graphs, often we use statistical modelling. Weekly sales can be analysed for trends, seasonality and autocorrelation, and to quantify special effects. A manufacturer of breathing equipment was intrigued to note a disturbing tendency for salespeople, whose sales rose steadily, to leave the company soon after – a clear case of competitors poaching star performers. Machine learning techniques, such as decision trees and random forests, can rationalise loyalty card data clearly showing the customer characteristics associated with higher spending, and guiding marketing staff to focus their attention on these customers.
Impact on everyday activities
The potential impact of simple data analytics on every day activities should not be overlooked! For example, a friend has a boutique and keeps records of how many customers enter the shop every day, what they buy and how much they spend. Looking at these figures over the first and subsequent years of trading we could see:
Weekly takings stayed steady over the first year of trading and then started to rise, which is good for a new business.
There were bumper sales periods around Christmas, Easter and the summer holidays, so we could predict how early extra stock should be put in place and when extra staff should be called in.
Sales per hour were steady, so that it was justified to have shorter opening hours on some days and longer on others and still maintain the income.
The owner achieved 50% more sales per customer than other staff. This could be because the owner is more motivated or skilled, but to be fair, it was probably because the owner chose more promising days to work.
From the company partner’s point of view, there are clear commercial benefits from data analytical projects. From the student’s point of view, it is rewarding to explore real data and apply all the graphical and analytical techniques they have learnt in their studies.
Service industries such as hairdressers, florists, taxis, heritage sites, theatres, art galleries, museums and healthcare can benefit from basic mathematical analysis of their operational data. If they are confident that their data will be kept safe and secure, companies are usually keen to collaborate and what fun projects these make.
The UK Knowledge Exchange Hub for Mathematical Sciences (KE Hub) is part of the UK’s national Mathematical Sciences infrastructure and exists to promote, facilitate, and support knowledge exchange activities between Academia and Business, Industry, & Government.
In October 2023, Shirley and Jo joined the KE Hub’s small team of part-time KE “Super Champions” to play leading roles in the KE Hub delivery.