On tomorrow’s doorstep - Innovation as a service, FAIR data and benefiting from failure and start-ups
CIONet’s September 28 event took place at the Ter Ham Castle in Steenokkerzeel, close to Belgium’s capital, Brussels. Let’s talk about the highlights and what they may mean for you.
What is FAIR data?Hans Constandt, founder and CEO of startup company ONTOFORCE, presented the audience with the limitations of current search engines. Suppose you would like to book a room in the Hilton hotel in Paris, and you type in ‘Paris Hilton’, you get something very different than what you had hoped for. Data may be the new oil, but like oil, it needs to pass through a refinery to get turned into useful products.
Mr. Constandt set out to create a platform that would include semantic context in searches, with pharmaceutical and medical sciences as its primary application. The vast amount of data in the platform has to meet the ‘FAIR’ standards, he says: findable, accessible, interoperable and reusable.
This means, in addition to having the proper contextual metadata, as many file formats as possible should be accessible from a technology-agnostic point of view, and the results should be able to be meaningfully linked for further research and analytics. Another key point of ONTOFORCE’s semantic search platform is that it does not involve a steep learning curve or requires extensive software development skills. On the contrary, it is a low coding platform and can be used by business people without extensive IT training.
Innovation in the power gridBut innovation isn’t just for start-ups. Eandis manages the distribution grids of electricity and natural gas in 239 Flemish municipalities and counts some 4,000 employees. Its Chief Innovation Officer, Daan Hostyn, spoke about innovative experiments happening at the company.
Their experiments are often done with technology partners like delaware and cover the entire range from incremental (step-by-step) to disruptive (immediately game-changing). They include IoT-enabled cathodic protection on the power grid, constantly measuring whether installations are sufficiently negatively charged to avoid metal corrosion; IoT-equipped sensors that keep track of humidity levels in electricity cabins, or detectors that send out alerts when cable reels are being moved unexpectedly, for example in a likely case of copper theft.
Eandis also runs AI test cases to see if asset lifecycle management can be improved by predictive analytics or whether cost savings can be found through RPA (robotized process automation) and is investigating uses of blockchain technology for certification management (e.g. green certificate trading).
Humanoid robots and cognitive servicesAt the conference, delaware explained how cognitive services can become easily available through the cloud (e.g. MS Azure) and demonstrated that robots seem to become more human every second. During the demonstration of these cloud services, our humanoid Pepper robot suddenly complained about painful joints and had to reboot itself to deal with a mechanical problem, much to the amusement of the audience.
As in the Eandis case, AI is not just used to replace manual processes but also to engage in predictions in ways that humans would never be able to do on their own. A good case was presented by pharma company UCB’s CIO, Herman De Prins. In addition to estimating the efficiency gains with RPA at 30 to 40%, a combination of machine learning with IBM’s Watson resulted in a 99% accuracy in diagnosing osteoporosis, even at very early levels where humans can’t detect any signs of the disease yet.
Failing isn’t bad, as long as you learn from itBoth delaware and Eandis’ Mr. Hostyn agreed on the key message that failing when experimenting isn’t a bad thing. Mr. Hostyn estimated that only around 1 in 10 experiments actually becomes successful, but that 8 in 10 failed experiments have their own kind of value. What really matters is that you learn from those.
And to be able to do so, big companies nowadays seek collaboration with nimble start-ups, which takes some learning of its own: it means adapting to the less-organized, no-frills mentality of a small team and keeping big, cumbersome frameworks and red tape out of the door.
Innovation is crucial, anyway. “Innovate or perish,” UCB’s Mr. De Prins called it. Named 2016’s CIO of the Year by Trends magazine, he confirmed the fact that big companies need to learn how to work with agile start-ups. Start-ups don’t have the same DNA as established companies, they don’t use rigid frameworks nor procedures and big companies have to review their supplier and risk management to be able to benefit from the collaboration. Indeed, the only thing that matters to an IT team is how much value they really create for the business.
Bottom-up or top-down innovation?Mr. De Prins is an evangelist for top-down ways of innovation – to overhaul an IT culture from the top, make quick decisions about budget allocations and direct people into new roles. 68% of the workforce find themselves in new roles, while 48% of all job descriptions within IT didn’t even exist 15 years ago. So, quick rotation among roles is necessary to remain fresh and keep learning.
At the other end of the spectrum is the bottom-up way of innovating, which is likely to generate ideas more organically and democratically, but with slower decision-making. Companies that attempt innovation will have to find out for themselves where they want to be in this spectrum.
Data and the future businessFinally and like ONTOFORCE’s Mr. Constandt, Mr. De Prins argued that value isn’t in processes, but in the data that you process. For Mr. De Prins, this is part of an evolution in the IT and the business landscape. Companies are evolving from classically functional, hierarchic organizations into a sum of projects and finally into a sum of products and services.
We at delaware are on a similar track. Innovation, we believe, can be offered as a service. By bringing together best and future practices, a knack for understanding new technologies and an intimate knowledge of our customers’ challenges, we are currently running innovation experiments with four different customers. These may or may not include a Pepper robot in good health, but they are definitely the areas where tomorrow’s businesses are taking shape.
If you’d like to read about an example where we go all-in on innovation, we have the Distriplus case study, where we are leveraging Big Data to create predictive intelligence and set up an omni-channel experience.