Site navigation

DataFest 2017. Cows, Killers and the Power of Data

Alex Leslie


You know a conference is going to have a certain buzz about it when it has its own bus. And Data Fest 2017, hosted by The Data Lab, had its own, very colourful, bus. And the event definitely had a buzz. As the writer and broadcaster Alex Bellos said, ‘Data Fest 2017 is another great Edinburgh festival’. It is safe to say that this jaded conference veteran was impressed.

The quality of speakers was extremely good, and very entertaining – a great achievement. Among the favourites was Mathematician and Broadcaster Hannah Fry of UCL (Twitter @FryRsquared – see what she did there) whose talk covered pedometers, cows, Harold Shipman and mosquitoes.

You might be wondering whether Fry was at the right conference, but the examples provided insights into how data allows us to do so much more in many different fields. The first field, if you will pardon the awful pun, is to be able to know when a cow goes into heat. Cows do not stay in heat for long, nor are there any obvious signs. What they do, though, is lie down, stand up, generally fidget. Thus the pedometer. Once you know the cow is in heat, data analysis has shown that if you have the bull scrubbed and sent up to the bedroom at a particular time during the heat, the chances of producing a cow (rather than a bull) are much higher. And cows, if you are a farmer, are a good thing.

Data also helped solve the Harold Shipman case. Shipman was a GP during the day, and a serial killer during the – afternoon – as it turns out. Shipman, who said at one point that he ‘had nothing to hide’, came really rather close to getting away with it. After all, GPs have patients who die. Not generally as many as Shipman, but they do. What proved very conclusive is that most GPs’ patients do not have a favourite time to die, while Shipman’s did. A huge majority of them died around 3 O’clock in the afternoon.

Another favourite was Jason Leitch, Clinical Director of NHS Scotland. He admitted early on that he had completely ignored the conference producer’s brief, and he talked about small data. What his talk on small data highlighted, though, was how making sense of data, large or small, is key to seeing the benefits. An elderly patient, John, had a heart problem and needed a defibrillator full time. He needed to get to Glasgow to get one fitted, but Glasgow was full. His heart surgeon had to stay with him in Ayrshire to make sure he stayed alive until he could get to Glasgow. John said he wanted to go home. His surgeon said that he would die. John said he wanted to go home.

There is a new(ish) thing that the NHS has introduced. When people are badly ill, they are now asked ‘what matters to you’? John was asked what mattered to him. He said, ‘my fuschias’. He was, it turned out, champion in the eight inch pot fuschia competition.

So the surgeon – as you do – sent a van to John’s house, picked up the fuschias and took them to the hospital. John was able to tend them. Once the defibrillator was fitted, John and his fuscias went home and he went on to win the competition for the third time.

There was an old lady with dementia who kept getting out of bed in the middle of the night and falling over. Finally, a niece came to visit and the nurses asked her whether she had any thoughts about why this might be. The answer was that she was desperate without her rosary. It was, they discovered, in the drawer next to her bed (apparently Health and Safety didn’t like the chances of keeping it clean) and when she had it back she was happy, and she stopped getting out of bed and falling over. And, oddly, people stopped by to talk to her, because she now had a story.

It is tempting to talk, non-stop, about the dark side of data. About Facebook plundering our data for ill-gotten gain, about Google getting into hot water for placing adverts within to racist videos, about how we all want our data back. But using data well is clearly a good, if not great, thing. Hannah Fry’s mosquito example was about identifying the whereabouts of stagnant water in Africa, by mapping malaria outbreaks. Identify stagnant water and you can eliminate it, and with it the risk.

As Frances Sneddon of Simul8 said, ‘we need to get people to trust us with their data, even though most of them have no idea how much data we share anyway’.

Edinburgh, it seems, is not just host to a very interesting conference about data, it is also a genuine hub for Data Science and Data Scientists. The Data Lab exists to encourage industry, public sector and university bodies to innovate and develop new data science capabilities and talent in Scotland. With over 40 funded innovation projects, 90 Masters students and community events across the country, it is clear that Scotland is becoming big business in data science.

Not surprising then that companies are even considering setting up here, even Silicon Valley ones. John Akred, CTO of Silicon Valley Data Science is one, although he admits it might also have something to do with our whisky. And getting hold of data to work with is not something that bothers Akred. He created an app in San Francisco that predicted how late trains were going to be by bombarding (surely politely asking, Ed) the train company via Twitter to provide it.

The hashtag throughout the event was #Datachangeseverything and the event proved this to be true beyond doubt. Let us just hope data changes everything for good.

Oh, and the last speaker, Costi Perricos of Deloitte, while talking about the hype and reality of Artificial Intelligence (our own view is that there is a lot that is artificial and not a lot yet that is intelligent) said the most interesting thing, ever.

‘Where, in the voice driven and controlled world of digital assistants, do you put the adverts’?

Alex Leslie

Founder - Disruptive Views

Latest News

Climate Transport
Editor's Picks Events Featured
Business Trending Articles
%d bloggers like this: