Search ST Engineering

  • Global
  • Global

Businesses today should invest in 4th industrial revolution (IR) technologies with a good sensing of the problems they wish to solve.  They should also understand that implementing technology and hiring data scientists alone are not the main ingredients to complete resolution to problems.  Businesses need to ask the right questions to get the right answers, and work towards transitioning management and staff into embracing and harnessing 4th IR technologies.  Speaking to more than 500 industry professionals at the annual ST Engineering Electronics Technology Seminar held on 27 June 2018 at Marina Bay Sands Convention Centre, Mr Bill Lee, Co-Founder and Managing Director of Azendian Solutions, a Singapore-based data analytics company suggested that businesses go into data analytics, only if they would be able to “make more money, save money or don’t lose money” from the investment.

This Technology Seminar on Artificial Intelligence (AI) is an exclusive customer event that brought together industry experts from the US, Israel, China and Singapore to discuss how AI is transforming the way businesses and governments deliver solutions and services.  It was organised to provide deeper understanding and insights for government and enterprise customers, reinforcing the need for businesses to relook, rethink and re-engineer themselves in the onslaught of technology disruptions.  ST Engineering has been strengthening its AI capabilities, having built its expertise in data analytics and working on projects over the past 10 years.

Offering the audience some business cases on how AI had been applied with business benefits, the Group’s President for Electronics sector, Mr Ravinder Singh, shared that in maritime surveillance, the use of AI meant some 2,000 ships passing through the Singapore Straits daily and 1,000 ships in harbour could be analysed for anomalies and deciding which particular ships could pose some security risks.  One taxi company also reduced the time it took to complete a taxi booking by half (from 26.9 seconds to 14.6 seconds) and increased the success booking take-up rate, by implementing data analytics and machine learning.  In addition, predictive analytics had helped train companies in China, SE Asia, Middle East and the US predict train door failures with 99.2% accuracy, seven days ahead before failure, in order to rectify issues early and avoid service disruptions.

Other applications of AI included the use of AI and machine-learning driven technologies like Alibaba Cloud’s ET Brain, said Dr. Derek Wang, Chief Cloud Architect, Alibaba Cloud International.  The ET Brain had been used to enhance traffic perception and traffic lights optimisation, proactively detect incidents, track hit-and-run and hazardous vehicles, and overall aid in public transport optimisation e.g. bus route and schedule optimisation.

The use of Smart Camera technologies, such as those from Yitu Technology, had also seen smart cameras play the role of being the “eyes of the city”, added Dr. Wu Shuang, Research Scientist at Yitu Technology.  Aside from facilitating traffic perception, with the help of suitable AI algorithms, smart shopping malls in the future could instantly detect the behaviours of shoppers and digitally push suitable offers to them real-time, even without knowing their identity.  By scanning a photo, police could also leverage smart cameras to quickly locate the whereabouts of a missing person.

Furthermore, fraudsters and suspicious behaviour could be detected with the help of AI, even if fraudsters take on multiple identities.  Using an AI process called “entity resolution” which involves matching different data sets, Mr Jeff Jonas, Founder & CEO of Senzing, had successfully applied AI to identify fraud in businesses and casinos, and even ships behaving suspiciously in country’s waters.

Yet the challenge on AI usage remains.  While AI is very real, and numerous applications exist out there, how do we properly train machines to learn something, and yet prevent it from learning the wrong things?  This is where AI research & development needs to continue, aside from providing the data that is needed to train machines properly, said Prof. Isaac Ben-Israel, Chairman, Israel Space Agency & Israel National Council for R&D.  Mr Jeff Jonas added, “Errors and diversity in data are good, and we shouldn’t throw them all away. You’ll lose your ability to find the signal. Systems that remember the errors are smarter.”

While the wheels continue turning for AI R&D, even whilst cities look to AI to address urban challenges in Smart Cities, can we look forward to a day when AI will solve all of mankind’s problems?  To Prof Ben-Israel, that is a difficult question and we are still years from figuring that out.


The Technology Seminar by ST Engineering’s Electronics sector was held at ConnecTechAsia 2018 in Singapore.