Chapter 9: Shipping

To anyone from an EU background, product liability traditionally evokes thoughts of strict liability under the Product Liability Directive of 1985. 2 This is of some relevance in the marine context, but (as will appear) only in a rather constricted way. It will briefly appear towards the end of this chapter, after the main part, which will concentrate on traditional product liability based on negligence.

In cases not concerning new technology, there is no serious doubt that on principle the general rules of negligence-based product liability apply at sea as they do elsewhere. 3 Thus shipbuilders whose negligence results in damage to other property can be sued by the owners of the latter provided the loss is not too remote and causation is otherwise established, 4 and the same goes for the manufacturers 5 and suppliers 6 of components. The availability of such a claim, moreover, takes on additional significance since it can give rise not only to a direct claim but also contribution proceedings. Hence the owners of a vessel found at fault in a collision claim can where necessary seek contribution from shipbuilders or component manufacturers in so far as the latter could have been sued by the original claimant. 7 Conversely, limitations on general product liability also apply, such as the bar on tort claims against manufacturers or suppliers for damage to the manufactured item itself. 8

These underlying principles are likely to survive any encroachment of digitisation (unless some international agreement can be reached, which seems unlikely). Hence except for claims for pure economic loss, and claims for damage to the ship from components fitted on board her at the time of acquisition by the claimant, suits based on negligence-based product liability will remain not only a possibility but a probability. But the practicalities are likely to change to some extent. One reason is that liability in negligence under the general principle in Donoghue v Stevenson 9 can be difficult to establish where there is a likelihood of intermediate examination or testing of a product. 10 This is likely to be highly relevant when it comes to any kind of electronic control system: most companies will not entrust the fate of $50 million worth of capesize bulk carrier or even $100,000 worth of containerised goods to such a system without pretty exhaustive testing and evaluation. Secondly, there may be other practical changes caused by the progressive replacement of people by code and EDI. This is likely to affect at least three fields in particular.

One is cargo documentation, identification and delivery. On principle this is ideal for digitisation. Repeated physical movements of paper over long distances are currently a nightmare for ship, cargo and financier alike; their supplanting by instantaneous transfer and reading of e-documents is an obviously progressive step. Admittedly attempts to digitise bills of lading, documents of title and other transport paper have hitherto had mixed success, and have not received anything like universal trust. 11 But the advent of effective blockchain and distributed ledger technology 12 may well change this by making such documents sufficiently reliable and tamper-proof to persuade shipowners to rely on them and P&I interests to provide cover, 13 and to convince congenitally sceptical financiers to lend on the basis of such documentation. 14 Similarly, the process of keeping track of containers – a major concern with container vessels laden with 20,000 TEUs now nothing out of the ordinary – has quietly gone digital. And whatever the teaching, so dear to law professors, about the need for goods to be released only against a paper bill of lading, it is worth noting that the trucker taking cargo away from the terminal will these days likely have obtained it by producing a PIN or other electronic authorisation. 15

Secondly, there is the care of cargo, as contrasted with that of the vessel itself. Aspects of this are already increasingly automated. Examples include continuous monitoring of temperature levels in reefer transport, and conversely the regular testing of the atmosphere around bulk cargoes liable to overheating or spontaneous combustion as a standard fire precaution. Similarly, for delicate cargoes like fruit, containers are increasingly more than boring metal boxes. They are often now decidedly smart devices, 16 with sophisticated contraptions to monitor, and sometimes indeed spontaneously to make changes to, the atmosphere within them on the fly. Yet again, stowage plans and their execution are now to a large extent automated, hence (one hopes) not only maintaining good trim and the logistics of loading and discharge, but also avoiding errors like the stowage of delicate or flammable cargoes next to engine-room bulkheads.

Thirdly, and much more in the future, there is the process of navigation itself; the day-to-day operations of getting under way, stopping, manoeuvring, docking, watchkeeping, routing, dealing with hazards and collision avoidance. Although the entirely unmanned cargo ship in commercial service is still very much a future project, 17 there have been successful experiments; 18 and there is already a definite prospect of increasing numbers of these processes being automated and crews being correspondingly reduced gradually towards skeleton numbers.

Will these matters cause product liability considerations to become more important in shipping litigation? In some cases, it must be admitted that the answer is often going to be no. In so far as the carrier’s seaworthiness obligation is concerned, for example, the carrier is already liable under the Hague and Hague-Visby Rules for the negligence of independent contractors to whom he entrusts measures to ensure it, 19 so the fact that he relied on an apparently competent supplier of electronic controls will not avail him if in fact the latter was negligent and as a result the vessel was unfit to carry the cargo safely. In such a case, third party liability will only be relevant if either the carrier is bankrupt and bereft of P&I cover, 20 or there is a need to escape limitation under the LLMC 1976.

On the other hand, in other areas the effect may be noticeable. Outside the field of unseaworthiness many other aspects of liability – for example, collision, injury to crewmen and many cases of liability under the terms of specific charters 21 – currently depend on ordinary vicarious liability for fault: that is, on whether employees have been negligent in the course of their employment. But in so far as employees are replaced by computers and algorithms, this solution is thrown out of balance. There is no such thing as a negligent computer; it follows that if owing to a computer malfunction something goes wrong the shipowner, unlike the employer of a negligent crewman, is liable only for personal fault (for example, a showing that they failed to supervise or operate the device properly, or had means of knowing that there was something wrong with it). Stress will therefore shift from the owner’s P&I club to those on shore responsible for constructing the system and the accompanying machinery, and those writing the code for it. Secondly, even where owners do remain liable, the involvement of shore-based producers may give further scope for third party or contribution proceedings by them against others who might have been held liable. 22 Third, it is also true that a growth in the scope of third party claims by cargo may provide an escape in collision cases from the “several liability” rule. Cargo claims against the non-carrying vessel arising out of an incident in which both vessels are at fault are available only to the extent of the proportionate fault of that vessel: 23 cargo claims against third parties are not so limited.

The fourth relevance of third party liability may be practically the most important, however. This is limitation, particularly relevant in the case of large claims against the owners of small ships. The LLMC 1976 allows shipowners to limit and extends this right to some people connected with them: the special case of salvors aside, these are the “owner, charterer, manager or operator” and their respective employees. 24 But that is the limit of protection. Except possibly where entire control of an autonomous vessel has been delegated (for example to an onshore control centre), in which case the delegate might be classed as an “operator”, none of these words seems applicable to a third party producer or provider of control services. It seems to follow that, absent any change to the LLMC 1976, product liability claims brought against such persons are not covered by it. But this is subject to the next paragraph: will courts be prepared to allow claimants to take advantage of this, or will they regard it as an illegitimate attempt to circumvent a wholesome regime?

The difficulty here...