Would Autonomous Ships Be Good for Society?
There has been a good deal of discussion recently suggesting that unmanned, autonomous ships represent the future of the maritime industry. The discussion has been primarily driven by EU-funded research that sees autonomous, unmanned ships as a key element for a competitive and sustainable European shipping industry in the future.
The EU has had a long-term goal of making short sea shipping more competitive with road and rail transport, which is under stress from the transportation bottlenecks caused by increasing volumes of internal trade. Faced with massive infrastructure costs to upgrade road and rail, the EU is making a concerted effort at the policy level to move large volumes of cargo from land to the “motorways of the sea.”
The discussion has also been influenced by ambitious press releases from equipment manufacturers and potential service suppliers, who are expanding the concept to include a proposed automated global shipping industry. This would create a new market for vendors, but it is not driven by user demand.
There is little doubt that advances in information and communications technology and robotics will impact the maritime industry and accelerate changes in the way ships are operated. We have already seen the effects of automation in the manufacturing and distribution sectors and the massive changes they have brought.
The concepts underlying autonomous ships are based on the “Industry 4.0” (1) model recently developed in Germany, in which cyber-physical systems would monitor sensors, create a virtual copy of the physical world and permit decentralized decisions. Its goal is managing automated production within a “Smart Factory” and the integration of multiple factories, suppliers, distributors and consumers through the internet of things or services. But it is questionable whether a model developed for a controlled manufacturing environment can be transplanted to the dynamic, global, uncontrolled and open maritime environment.
Given both the unique nature of the maritime industry and the exponential growth in the application of technology, any attempt to predict the extent and consequences of automation is speculative at best. Any change will come in stages, and each stage will require evaluation.
One stage will certainly include remote monitoring of all functions aboard ships with far greater shoreside management of shipboard operations. To some extent, for certain functions, remote monitoring has been around for decades. Cost considerations and bandwidth have been limiting factors, but now that is changing. Efforts at the International Maritime Organization (IMO) to develop and implement technology to support e-Navigation – including interoperability and harmonization of information between ship and shore – could potentially be a precursor to remotely controlled ships. That raises, at least as a concept, the possibility of truly autonomous, unmanned ships making their own decisions using artificial intelligence. While this seems like fantasy at the present time, given the increasing capabilities of technology it cannot be completely ruled out as a possibility in the decades ahead.
The pace and extent of how this will all evolve in international shipping will primarily be determined by economics and risk factors. It is anticipated that the cost of building a ship with the required technology and redundancies for remotely controlled operation may be higher than that of a conventional ship, even with the elimination of the crew’s accommodations. The system would also require shoreside infrastructure with a global reach for monitoring and control, as well as expensive shoreside support for maintenance, repairs and functions now carried out on conventional ships by seafarers from relatively low-cost-labor-supply countries.
Can the additional costs of an autonomous system be offset by substantial reductions in crewing costs, or their complete elimination? If not, there is no economic justification for ship owners to shift to autonomous ships.
It is estimated that crewing costs are only about six percent of the overall cost of running a ship. (2) Capital costs are about 42 percent and voyage costs, including bunkers, run about 40 percent. There can be no doubt that shipping is capital intensive rather than labor intensive.
Even if all manning were to be eliminated, it is not at all clear how the relatively minor savings in crewing expenses could compensate for the additional costs of building and operating a remotely controlled autonomous ship system and its supporting shore side infrastructure.
There has been a disturbing degree of enthusiasm in some circles for autonomous ships, which would create massive unemployment of the world’s seafarers and disrupt the economy of the maritime labor supply countries, all to achieve a rather minor reduction in the cost of shipping. But so far there has been a noticeable lack of enthusiasm for autonomous ships among shipowners: perhaps they have a better grasp of the economics than the enthusiastic proponents.
Maersk has indicated that it may be looking at some form of autonomous ships in the 2030-35 timeframe, which would coincide with the end of the useful life of their recent newbuilds. But ship owners will only embrace autonomous shipping if it is commercially viable and they can gain a competitive advantage by eliminating the costs of seafarers. As we see from the statistics, however, the extent of any cost reduction accomplished through cuts in manning would be limited at best.
Advocates of autonomous ships are attempting to justify their position based on a shortage and a fictional lack of competency of seafarers. They propose a “solution” they say would lower costs and increase safety through eliminating seafarers and with them the risk of human error. They fail however to acknowledge the very real risk of introducing new sources of error in technical systems, communication links, cyber security and remote human controllers who are isolated from the reality of the ship and its actual environment: the greater the complexity, interdependencies and links within a system, the greater the opportunity for errors and failures.
We should be skeptical of optimistic projections of the future benefits and efficiencies of autonomous ships that are based on the aspirational views of advocates who have a commercial stake in creating a market for an autonomous ship system. A great deal of investors’ money has been lost in the past through bets on over-hyped expectations regarding new technology. The “Dot-com” bubble of 2000 is a good example.
There is also a misconception that complex, highly automated systems require highly skilled operators. But neither the experience of other industries nor the academic literature support such a view. As the technology assumes increasingly complex functions, there is a down-skilling of operators who become dependent on highly automated, self-regulating systems. The more automated the system, the less is required of the human operator since basic competencies and lower level decision making functions are built into the technology.
The human operator loses the opportunity to develop through experience the deeper skills and talents, such as assessment and judgment. (3) The problem with machines that think is that they give rise to people who don’t need to think. This presents a significant challenge in the dynamic and complex maritime world, where assessment and judgment, based on experience and total situational awareness, are fundamental to making the “right” decisions – often under tight time constraints – that can spell the difference between a safe passage and disaster.
Replacing skills and active participation with the boring task of monitoring displays can lead to a dangerous level of complacency. When things go wrong, the handover problem between the automated system and the complacent, uninvolved human operator—with degraded skills and situational awareness—has proven to be a major problem. The significance of this problem has been demonstrated in the air transport sector and other industries.
There is, in addition, the need to take a much broader view of automation, beyond its impact on the maritime sector. What is required is a realistic view of automation’s social and political consequences in the world at large. The assertion that technology is a purely benevolent force, whose only impact on society is a positive one, is clearly erroneous. There is no economic law that says that everyone, or even most people, will inevitably benefit from technological progress. Technology is neither good nor bad. Unlike humans, it has no moral or ethical values.
Technology is a powerful force that can destabilize institutions and industries, upset the social contract between capital and labor – as well as the traditional employer/employee relationship – with profound consequences. The changes it brings have been compared to a fourth industrial revolution. (4) And there is considerable concern in academic circles regarding technology’s impact on the future of our society, on our economic system and on our political institutions. The issue is not solely of concern to people being replaced by technology; this year at the World Economic Forum at Davos, it was the main topic of discussion among global leaders.
Academics estimate that as much as 50-70 percent of the labor force can ultimately be replaced by technology. The gains that automation produces from increased productivity with lower labor costs primarily benefit the capital investors who own and control the technology. The resulting profit-concentrating effects, coupled with technological unemployment or under-employment, are largely responsible for the increasing inequality of income that is creating social tensions and political turmoil in the United States and elsewhere. Globalization has been the ready scapegoat of politicians, but technological unemployment is responsible for much of the problem.
Technology-generated income inequality is also a threat to our free market economic system because it reduces consumer demand by reducing the number of consumers, principally workers, with the ability to purchase goods. Our free market system is based on consumer demand driving manufacturing and production. In past industrial revolutions (steam, mass production, electricity) an increase in demand created more jobs in production. Those jobs would in turn increase consumer spending and drive up production. In the fourth industrial revolution, with technology replacing workers, this classic economic principle no longer holds true. The cycle between consumer demand, production and jobs has been disrupted. Automation now allows for a scaling-up of production coupled with little if any increase in the number of jobs. This is evident in the economic data: corporate profits and the stock market go up, while worker income and consumer demand are near stagnant.
The factors of primary concern to society are: the consequences of the change in the nature of work (or even its elimination); the distribution between capital and labor of the enormous gains in productivity that derive from technology; the preservation of the traditional employer/employee relationship, which provides stability and social benefits; and the effect on the future of society as a whole of technology and robotics replacing or down-skilling workers.
In a globalized industry, these are not issues that lend themselves to easy solution. They involve economic and political issues – which impact the society as a whole – regarding the distribution of productivity gains derived from technology. In our democratic free market system, individual economic decisions are determined by self-interest. Matters of common interest to the welfare of society at large should be addressed on a political level through good governance. The negative consequences of the disruptive power of technology may not be the fault of technology, but the failure of government policies to address its impact on society.
Where technology ultimately takes us will be decided within political institutions, legislative bodies, regulatory agencies and international organizations such as the United Nations and the IMO. What is needed is a common understanding of the issues and a coordinated effort to protect not only the interests of seafarers and workers but also the future of our society as a whole.
These thoughts have touched primarily on the broader economic and social issues. There is a great deal more to be said on the technical aspects—legal and regulatory, ultimate responsibility and liability for risks, human factors and man/machine interface issues, software quality, cyber security, reliability of the communication/data links, and engineering, sensor and technical systems—that will be left for another discussion.
Capt. George Quick is Vice President of the Pilot Membership Group at the International Organization of Masters, Mates & Pilots (MM&P). He serves on the delegation of the International Transport Workers Federation (ITF) to the IMO Maritime Safety Committee, the IMO Legal Committee and the IMO Facilitation Committee.
(1) "Industrie 4.0" originated in a working group organized by the German government to promote the computerization of manufacturing. The final report of Working Group Industry 4.0 was presented in April 2013.
(2) Martin Stopford, “Marine Economics (Third Edition),” page 225. Based on data for a Capesize bulker. At today’s bunker rates, my rough calculations are that manning costs fluctuate between 3% and 5%, depending on the volatility of bunker rates and their impact on voyage costs.
(3) “The Glass Cage, Automation and Us” by Nicholas Carr.
(4) “The Fourth Industrial Revolution” by Klaus Schwab, Executive Chairman, World Economic Forum, Davos.
The opinions expressed herein are the author's and not necessarily those of The Maritime Executive.