The automated container terminal and the odd (mis)adventure of the Google car


At a recent dinner with friends, the topic of our conversation moved to the invincibility of autonomous vehicles and the unfortunate encounter between the bus and the Google car. It was my turn to be ‘the heart of the party’. Score one for rational thinking and mathematics.

Fans of autonomous flying vehicles (a.k.a. drones), proponents of self-landing planes, enthusiasts of autonomous driving vehicles, and whatever other sorts of unmanned flying, sailing and flying things of the future – it’s time to pause your joysticks for a refresher course on mathematics.

Can we predict the unpredictable?
Even the best designed objective function cannot stop unpredictable human behavior. From the street, where a bus driver does not yield to a righteous mathematically-driven Google self-driving car, to the container shipping terminal allowing driverless prime movers, rubber-tired gantries (RTGs) and straddle carriers to mix with manned trucks, we need to urgently solve one thing: Either we forget autonomous vehicles altogether, or automate everything and let no human drive. Pretending we can control the mixed ‘human-machine’ environment and expect no collisions is insane.

To the enthusiasts of ‘salvation by big data’ and the Internet of Things (IoT) – let’s consider the facts. The Google vehicle has collected 3,500,000 miles worth of driving data through a reasonably sophisticated array of IoTs, a massive amount by any stretch of imagination. And only to conclude, in 2016, that bigger vehicles (driven by humans or otherwise) may not yield! OMG! A Google programmer just learned the laws of physics – that the braking distance of an object is proportional to the mass and velocity of the said object! Since those objects cannot stop on the dime, other vehicles are strongly advised to yield, even if they have the right not to. Maybe Google can hire a few railway engineers to help with that understanding. The study of physics and human behavior would be very useful for those programmers, no doubt.

What automation can and cannot do
Where am I going with this post? To ridicule Google programmers? No, not at all. I want to go straight into the heart of the fully automated container shipping terminal. I am not really interested in Google vehicle having reportedly 70+ accidents with other vehicles before the critical moment in 2016. I am only interested in this event, because it shows us the future, or limitations, of an automated container terminal.

In 2011, the Maritime and Port Authority of Singapore (MPA) and PSA Singapore Terminals, operating one of the world’s largest container transhipment ports, pumped in S$50 million each towards R&D in port technologies.

One of the most interesting projects under that valiant initiative was the development of the driverless automated guided vehicles (AGV) for container terminals. Why? Well, here is a bit of mathematics for you. The new PSA terminal was envisioned to move nearly 70 million containers per year, about double of the 2015 volume, while facing the simple truth – with the rapidly shrinking Singaporean workforce, there are less and less people willing to work, rain or shine, in a container yard. Even the expats decided to give it a pass! Apparently, even banking was more interesting than moving steel boxes around the yard and playing with really powerful lifting machines. As a result, PSA needs to automate or forget the envisioned growth.

Complete or selective automation?
Even if the terminal did not face workforce shortage, I honestly think that the dream of having a mix of human-operated machines and mathematics-operated machines inside a tight, confined space of a container terminal will not work perfectly. Until we achieve 100% certainty in actions/reactions of mathematically-driven/flown/sailed devices, my firm belief is that we have to decide between two propositions: Either have all equipment and operations automated 100%, or allow some automation here and there with built-in rubber bumpers all around. Simply because there is no way that rational mathematics can be programmed to avoid any type of irrational human behavior.

Can I say that my prophecy means the end of Singapore PSA as the largest container transshipment hub in the world? It could be. There are more abundant, and still cheaper, labor pools in Malaysia and Indonesia. There are existing and upcoming investments in deep sea ports in the region. In fact, with the addition of the high-speed freight link from Kunming (China) to the ports of Malaysia and Thailand, the existing ports in the Malay Peninsula will eventually replace Singapore as the ‘must stop and tranship here’ place. Even if PSA manages 60 moves per hour, it will not be sufficient to save the Singaporean terminal from losing significance, unless…

Unless, of course, mathematics comes to the rescue. The entire hope for Singapore to become the hub and center of innovation for the 100% unmanned container terminal, rests on the assumption that the kids studying mathematics in Singapore are much smarter that than programmers at Google. So, to the students of mathematics in Southeast Asia, here’s my message: Forget the ‘Go West’ slogan. There’s gold in mathematics – you just need to dig for it. The future of container shipping port in Singapore, and many other places, is in your hands and minds!

If you found this topic interesting, leave your comment or suggestion. I look forward to responding to your feedback.

This post was previously published on LinkedIn.