SINGAPORE: Nawal Roy is only too eager to rattle off strategies over a map of Singapore. Marked with icons and numbers, it represents a summary of his company’s efforts in Singapore's war on diabetes.
But unlike the doctors and healthcare professionals who are their comrades-in-arms, his team at medtech start-up Holmusk fights the battle on the third floor of an open-concept office block. Their weapon of choice - a wealth of aggregated and anonymised clinical data, all culled and analysed from 32 hospitals in the United States, together with local reports from a partnership with healthcare provider SingHealth.
Diabetic users of Roy's app, GlycoLeap, can tap this rich database simply by taking a photo of their lunch or sharing the results of their latest workout session. In return, the app provides individually tailored advice to help diabetes sufferers better manage their condition. Such advice would only get more effective with a larger, diverse set of data, said Roy.
“The more information you can capture and use, (the more you’re able to) create meaningful content for users or stakeholders, like doctors - or the government, for population health-related policies,” he said. “All that becomes powerful in the sense of how many data points you are able to capture.”
Holmusk is one of just a handful in Singapore that actively shares data with external parties, despite widespread recognition of data and analytics as the "new oil" that powers the global economy – an industry that will grow from US$130 billion (S$182 billion) in 2016 to US$203 billion by 2020, according to predictions by market research firm IDC.
In the recent Committee of Supply debates, Minister of State for Communications and Information Janil Puthucheary noted that many firms still lacked awareness of the possibilities of big data, did not have the expertise, or had concerns over regulations such as the Personal Data Protection Act. This prompted the Government to establish the Data Innovation Programme Office (DIPO) to promote greater data exchange between firms.
But Roy is firm that the private sector will need to take the lead to build a more effective healthcare sector. He cited overseas efforts, where data-sharing between Swedish companies and the government has led to advances in cardiovascular research, as well as developments in the field of predictive medicine in the United States.
“It has to be at the business-to-business level. The government can only be an initial facilitator - setting a perspective, or setting a regulation by which this (sharing) behaviour can happen,” he said.
“If the market is dominated by multiple players, then there’s an advantage for each player – to literally contribute, but also in exchange, get information that they don’t have, in meaningful values for each of them. I strongly believe that it is prime time to make data science as a key component of healthcare in the overall decision-making.”
But according to Deloitte public sector leader Lee Chew Chiat, many firms remain concerned about the competition figuring out their strengths and weaknesses. Others wonder whether there are any tangible benefits in return for sharing.
“Let’s look at the companies in largely two different groups. The start-ups will have no inhibitions whatsoever. They will see the opportunities in sharing, in doing things differently,” said Lee. “Then you get the incumbents who are asset-heavy. Their starting point would be: ‘What is the risk of sharing; what are the implications of sharing?’ So for this latter group of companies, they don’t have that natural intuition to want to share, because their starting point is risk and exposure.”
“So then the question is for Singapore – how do we enable companies to then overcome this inertia?”
Nawal Roy (L) pores over data collected from GlycoLeap. (Photo: Loke Kok Fai)
DATA-SHARING PIONEERS IN THE BANKING AND FINANCE SECTOR
Managing risks and the exposure of sensitive information is par for the course for Singapore’s banking and finance sector. And banks such as DBS and OCBC have been among the first to explore the possibilities of data-sharing across sectors.
Both have started and driven initiatives with the retail and telecommunication sectors, helping to improve targeted marketing efforts, such as catering to the taste of a target demographic and customer service, among others.
“We’re still experimenting with data-sharing. It’s quite sensitive with all of the personal data protection and our own need to keep our client confidentiality high,” said DBS Bank’s head of group technology and operations, chief information officer David Gledhill. But he added that aggregated data – collected from a group and stripped of identifying information – offers some freedom.
“For example, we’re starting to experiment with aggregated flow through stores and department stores and shopping centres. We start to have the information through the flow of card usage, (tying it to) for example, where do you want to put the next Uniqlo,” he said.
“We have flow data, we know shop-by-shop, location-by-location where customers are, what the footfall is, what the demographics are.”
But such efforts are not without heavy demands on resources. According to OCBC, such bank-driven initiatives can cost tens of thousands of dollars, requiring months for legal teams to work out and address legal issues. It added that while the banks might have the resources and expertise to do so, few other players in the other sectors could afford to.
“Basically, if I wanted to have a sharing arrangement with another company, we have to have a bilateral discussion, we need to find a place where we can load the data, draft independent legal contracts – it’s a lot of time, it’s a lot of effort, and everything is on a bespoke basis,” said OCBC’s head of group customer analytics and decisioning, Donald MacDonald.
ROAD BUMPS TO GREATER SHARING
According to OCBC and other industry players, part of the difficulty lies in finding a trusted middleman - a neutral party or platform accessible to all players.
Besides having the hardware to store collected data and the skilled manpower to analyse it, such an entity would need to set standards for effective data-sharing. It would also need to put in place a framework to negotiate legal concerns, privacy issues and cyber risks. Most importantly, it needs the trust of all industry players, the authorities and the wider public that the data it manages and holds would not be misused.
Energy start-up Green Koncepts could be seen as one such private sector entity. It developed and now runs a cloud analytics platform, crunching energy-use data from the buildings of clients to help them gauge their power consumption. By trusting the company to handle their corporate data appropriately, its clients reap the benefits of being able to benchmark their energy use against aggregated data pooled from other clients.
Green Koncepts founder and CEO Kenneth Lee said such an arrangement is superior to the open sharing of data, with agreements and terms in place to spell out how shared data will be used.
“There have to be some benefits that go back to the party that shares the data. And this has to come in the form of data analytics applications, which can solve their real-world problems. And I think we are in the stage that these applications are just beginning to be built, and the mass adoption of these applications has not actually happened,” said Lee.
Green Koncept's KEM Platform allows its clients to benchmark their energy use against others. (Photo: Loke Kok Fai)
The other challenge lies in quantifying just how much each type and unit of data is worth, to ensure a fair trade. With the space still in its infancy, industry players say there’s still no real consensus on data value, nor a standard to measure it by.
But some like DBS have already taken small steps towards establishing the economic value of such insights. And larger technology players are also doing much to open the space, said Gledhill.
“If you look at, for example, the things that Google are doing, they’re starting to give (out) aggregated data. IBM - through their whole cognitive approach - are starting to aggregate sources of data which they will package, bundle and sell on as application programme interface services et cetera," he said.
“So you can start to see it accelerating very, very quickly. I think a lot of these technology vendors will ultimately be the source, collection points and distribution points over and above what the government is doing.”
And with greater sharing especially across sectors, analysts said the value of each dataset could be maximised for both companies and individuals. Deloitte’s Lee gave the example of Japan, where a consortium of companies have come together with government help to create a trusted databank in certain sectors such as healthcare.
“From a consumer standpoint, you actually gain from it. You benefit from maybe lower prices, better engagement or more options,” he said.
“Now the value is that once you look at it cross-sectorally, the value is optimised. It is also hard to replicate - for another country or another place to try to replicate that, it’s actually very hard. So I think that’s the value the Singapore government sees.”