Connecting Dark Assets: An ongoing series on how the Internet of Everything is transforming the ways in which we live, work, play, and learn.

The Internet of Everything (IoE) is all about connecting people, process, data, and things in innovative ways. And if you think that’s just a nice vision for the future, look no further than the Hong Kong subway system — preferably after midnight.

That is when the last train rolls off the tracks and an army of about 10,000 maintenance people spring into action with their human actions carefully coordinated by a non-human intelligence. An algorithm-driven artificial intelligence (AI) system monitors the entire subway line to determine critical maintenance tasks and sends crews out accordingly. A damaged track in the north of Hong Kong, for example, will take precedence over a routine smoothing of the rails elsewhere.

Of course, if the AI system overlooks a problem, humans can override it. But Hong Kong’s 99.9 percent on-time record is a testament to the system’s high efficiency.

The MTR Corp., which created the AI system, is planning to roll it out in Beijing and other cities around the world. In the meantime, the Hong Kong system reveals people, process, data, and things in action. Sensors along the many miles of tracks, switches, and signals (the things) generate gigabytes of data that are analyzed in real time, generating actionable insights and highly informed decision making. Then the people and process side of IoE enters the picture.

The maintenance workers know exactly where to go and when to do the work without spending time reviewing planning schedules or discussing priorities. The automatic system even monitors noise levels to ensure compliance with local rules about nighttime disturbances.

It’s a great example of machines and people working together harmoniously. In the new world of IoE, the ultimate beneficiaries are humans. In this case, it’s subway workers and subway riders alike.



Joseph M. Bradley

Global Vice President

Digital & IoT Advanced Services