URSABLOG: Technologically Challenged

Maybe I have been a bit distracted of late by ship sales, travelling, education and Donald Trump but it’s been quite a long time since I’ve teased the tecchies. The steady drip-drip of ridiculous notions has now accumulated in my “Tecchie” folder to such an extent that I can’t hold back my scorn any longer. Two separate technologies are in my spotlight today: blockchain and autonomous ships.

Blockchain is the technology behind Bitcoin, and the hype behind Bitcoin is still going strong. I first really started noticing it over the summer when the nephew of a friend of mine started explaining how this was going to change the world. Then another friend of mine made a killing as an early investor; he cashed out and then used the profits to invest in high powered computers to mine for more. It is, by all accounts, slow work, and the electricity bills make the eyes water.

There was to me – and as the FT pointed out in an article in their Someone Is Wrong on the Internet series recently – an “inherent and fatal contradiction of something trying to be both a speculative asset and a currency useful for buying wheelbarrows, lattes or running shoes. No-one uses gold bars from the safe for petty cash.” People were buying it to sell it, always a risky game; just ask shipowners who bought ships in 2007 and 2008 for that very purpose.

Ah, say the blockchain evangelists, you may be right but the technology behind it is cool. Even our old friend Yanis Varoufakis weighed in: “Bitcoin is the perfect bubble, but blockchain is a remarkable solution.” Yes, but when Yanis (it still pains me to write his name) starts recommending something as a remarkable solution, we should be afraid, very afraid, especially when it involves money.

Blockchain, for those that don’t know is basically a shared ledger of a continually growing list of records, or blocks, which are linked and secured using codes or cryptography so that they cannot be modified. It records transactions between two parties in a verifiable and permanent way, usually over a peer-to-peer network. If you want to change the data in any given block, then all subsequent blocks will be altered, so the network will all find out and have to agree with it. Therefore it is the  ultimate trustworthy system of exchange. Clear? No, probably not. Let me try and put it another way.

A member of the network requests a transaction, let’s say to buy cryptocurrency or virtual coins. This request is broadcast to the members of the network using nodes (computers to you and I) and the transaction and the user’s status is then validated by algorithms. (This transaction needn’t be cryptocurrency, but could be contracts,  or even bills of lading and documentary letters of credit.) Once the transaction is verified it is combined with the records of all the other transactions to create a new block of data which is then added to the chain, the blockchain, which is permanent and cannot be altered retrospectively.

The powerful idea behind blockchain is that the rules are enforced mathematically so you don’t need a government or enforcement agency to come in and tell you that your transaction is false, or bully you, or insist on bribes to make the problems go away. But this selling point of this technology hides a few problems, at least for me.

Firstly, I think I get it, but I’m still a little unsure how it works under all conditions if I am perfectly honest. I’ve had a lot of people explain it to me, but I’m not sure that I’ve got all the pieces in the right place. Secondly, it is sold on the idea of trust, but it’s a very cold and mathematical trust, i.e. that the data, the records, cannot be changed. What if the data is not right, or even fraudulent, to start off with? The absence of human gatekeepers makes sure that trust between humans, and other old-fashioned words like integrity, honesty and so on, are forgotten.

Imagine bills of lading exchanged via blockchain (like the the big liner companies are pushing at the moment). Will it really mean that people will check cargo properly, and refuse to sign cargo that is not in good condition or of the right quantity? I suspect not. The boxes will be ticked, and the carriers will be liable, all in the name of efficiency. P&I clubs will develop new departments. And remember that bills of lading are not just used to record the receipt of cargo on board, they trigger payment under documentary letters of credit and other methods of payment, and also give possessory right to receivers. Can all this really be done without human oversight? And if it can, is it desirable?

Which brings me to autonomous ships, or as a recent article in vice.com called them, boats. The gist of this inadvertently funny article was that as humans are the cause of most accidents at sea then we need to replace them with machines. Michael Johnson, founder of Sea Machines, starts off by saying (using the example of Costa Concordia):

Ships worth hundreds of millions of dollars shouldn’t be able to be manually driven onto the rocks. We have the technology available to control these vessels.

He fails to point out the technology already exists to control the vessel in question and was actually on board at the time. The captain chose to override it for reasons of his own. Anyway, ignoring this important factor, he is currently testing a connected, advanced driver-assistance system for workboats. As far as I can see, it’s a remote control autopilot thing, and not really the kind of thing that could be applied to larger ships. And lest we forget, humans can make errors with joysticks as well as on board ships. It doesn’t seem like a solution, but just a sales angle for some cool new kit.

This visionary article goes on to say:

In the autonomous revolution that is underway, nearly every transportation machine will eventually be self-driving. For cars, it’s likely going to take decades before we see them operating freely, outside of test conditions. Some unmanned watercraft, on the other hand, may be at sea commercially before 2020.  That’s partly because automating all ships could generate a ridiculous amount of revenue.

Then they do a quick bit of Googling and come up with the following statistics:

According to the United Nations, 90 percent of the world’s trade is carried by sea and 10.3 billion tons of products were shipped in 2016. According to NOAA’s National Ocean Service, ships transported $1.5 trillion worth of cargo through US ports in 2016. The world’s 325 or so deep-sea shipping companies have a combined revenue of $10 billion.

Only “325 or so deep-sea shipping companies” in the world? There’s more than that in Greece! And $10 billion doesn’t sound very much to me. Anyway, they go on:

 As it turns out, autonomous systems for boats aren’t supremely different than those of cars, beyond a few key factors—for instance, water is always moving while roads are not, and ships need at least a couple miles to redirect.

 These are indeed key factors, and hard to correct once you get them wrong. But not to worry, there are great brains on the case. Thiru Vikram, CEO of Buffalo Automation, which has a new AutoMate system (sensors and cameras) made the following amazing discovery:

Through testing in the Great Lakes, Vikram discovered that only one ship at a time can pass through its locks and canals.

Thank God for that. He points out:

If a boat arrives early or late, it may idle and waste fuel. Algorithms and sensors could be used to interpret weather and seafaring conditions, and automatically adjust speed to meet the scheduled time, he said.

What does he think happens now, without algorithms and sensors? Has he never heard of AIS? Or agents? The article goes on:

At the same time, locks and canals are also narrow and congested, increasing chances of hitting another vessel or running aground. These risks of allisions (when a boat hits a stationary object) often mean higher insurance expenses, said Vikram. Automation, he continued, can reduce risks and costs.

Without saying exactly how. It’s probably a trade secret. But don’t worry, they have seen the future. Michael Johnson is confident that we in shipping, and the regulators, will see the wisdom of his vision:

 “Every company has their mandate that safety is first,” he said. “But definitely, things that greatly improve the bottom line do see pretty rapid adoption.”

No Michael, most shipping companies have survival as their mandate. Survival of their ships, their crew, and their finances, if the sea and the markets let them. They use their judgement to guide what tools they use, not delegate their judgement, whether in the office or on the bridge, to “technology”.

The story of the Costa Concordia disaster is a very human one, full of arrogance, sex, hubris and their consequences. Technology tends to enhance and multiply these factors, not eradicate them. And we are talking about ships here, not boats. I know of no owner that would be comfortable with completely unmanned ships, because they deal in another currency: respect. Respect of their crew, respect of their ships, and more than anything, respect of the sea.

Technology is after all only a tool, a tool in the hands of fallible human beings. How it is used, for good or ill, is up to us. I am not against technology at all, how can I be, if technology is after all the practical application of knowledge and experience? But I do not think the solution to these problems is getting rid of humans, but training and education in the widest sense of the words, to encompass not just the practical, but the ethical, having faith in the human brain and in human judgment, and investing in them.

We are the keepers of our destinies, and we should never delegate that responsibility to machines. After all they are only doing what they were designed to do, and we designed them. Blockchain tries to eradicate the problem of fraud, human error and a lack of trust by eradicating the involvement of human beings. Autonomous ships will solve the problem of human error by getting rid of humans on board ships. We should remember that power without responsibility is a very dangerous weapon, and wielding it very rarely results in the optimum outcome. 

Simon Ward

www.ursashipbrokers.com