SAP chooses Singapore for its global Innovation Centre for Machine Learning: Page 2 of 2
By Zafar Anjum January 2, 2017
We are not trying to build self-driving cars but self-driving enterprises
According to Christian Boos, VP, SAP Innovation Centre Singapore, SAP is not trying to build self-driving cars (like Google or Uber), but helping create self-driving enterprises.
In the startup-like environment, we explore unconventional ideas and develop inspiring proof of concepts, said Boos. The Innovation Center Network, which focuses on exploring new concepts and aiming for success in new markets, has attracted many co-innovation partners wanting to work on disruptive business ideas. We follow the Design Thinking methodology, and the users of our software are at the heart of everything we do, he said.
“We have 25 industries and we have industry experts around the globe,” added Pecht-Seibert. “We develop our solutions in Co-innovation mode. We do the pitch to our industries and they get the idea what Machine Learning can contribute and they come up with their insights into markets. Sometimes customers reach out to us with specific problems. That’s how we interact with each other and prioritize a market perspective and a business need perspective on the applications stack we plan to develop.”
In the last few years, SAP has worked on many Machine Learning projects that have become success stories, be it in the domain of academics, enterprise customers or government, Lee said.
For example, SAP was part of the National Science Experiment (NSE) – a large-scale experiment to help young citizens (students) understand science and technology and also to make the city smart (part of the Smart Nation initiative). It was an island-wide outdoor science experiment carried out by Singapore students to track their carbon footprint, travel mobility patterns, and the amount of time they spend indoors and outdoors.
Under this programme, selected school children were given a specially designed sensor to carry along with them for a week, meant to collect data on their daily travel as well as data from the environment.
This data was transferred wirelessly to a central online portal from which these students (and their teachers) could log in to view the results, including the aggregated data of students from all over Singapore who are taking part.
Through this experiment, students learnt about the Internet of Things and Big Data. Conversely, the data helped track the lifestyle of students: how much of their lifestyle was sedentary, how much exercise they got, and how much carbon they emitted on a daily basis.
All this could be measured and contextualized through data analytics. SAP provided help through its big data platform, SAP HANA, and their data scientists helped in data analysis.
Similarly, the SAP lab helped the Maritime Port Authority of Singapore in better port management through the ‘Maritime Intelligence for Port Ecosystem’.
SAP and the Maritime Port Authority of Singapore collaborated in the area of maritime intelligence for creating smart harbours. SAP Research has built a prototype of a software system that specifically uses Automatic Identification System, SAP HANA and Complex Event Processing engines for detection and reporting of maritime complex events (anomalies).
The objective is to explore new paradigms in maritime intelligence for safety and compliance to improve the productivity of port ecosystems. The specific use cases of anomaly detection that are implemented in the proposed software system are:
- Detection of ships that do not have proper flag information;
- Detection of ships that broadcast ambiguous identification;
- Detection of ships that engage in brushing incidents for longer periods of time; and
- Detection of ships that leave/enter particular areas (zones) as defined by the users.
But the most remarkable success story, according to Lee, has been in medicine. Using Machine Learning, SAP’s team of 10 in Singapore provided ‘Medical research insights’ (MRI) that helped oncologists to come up with personalized treatment for cancer patients.
“Survival rate of cancer patients was low because doctors followed a common process like chemotherapy, etc., for all patients. Now they can customize or personalize the treatment and chances of survival of patients are higher,” said Lee.
The Singapore team worked with the main research team in Postdam, Germany, for a year. Their contribution was critical, he said.
The research project was so successful that MRI scaled to a global platform and became a new business unit (called Connected Healthcare) within SAP –led by Steve Singh, who sits on the executive board of SAP.
“Work coming out of here (Singapore) was the foundation of a new business unit, and we got a design award too, the Red Dot Award for most user friendly and impactful application,” said Lee.
“Machine Learning is a very strategic topic for SAP, and has the potential to making all our enterprise applications more intelligent,” said Pecht-Seibert.
“It would help our customers to dramatically improve efficiency in their domains such as accounts matching payments against invoices which is typically a manual process today and release their resources to perform higher value tasks. That’s our way forward. We start with our applications and we learn and apply Machine Learning algorithms to real-life problems and derive insights from them. We want to develop a Machine Learning platform for our enterprise applications that will be available to our ecosystem partners and customers, and even to go beyond the existing applications.”
(Zafar Anjum is DNA's Contributing Editor in Singapore)
Are data analytics companies able to serve the ‘multi-faced’ Asian customers?
Amazon Web Services trains its spotlight on artificial intelligence
Microsoft CEO Satya Nadella outlines plans to democratise AI
SAP to invest €2bil in IoT by 2020