Conferences
Reporting back from conferences, seminars and discussions across the maritime
Î world. Join the discussion at LinkedIn, or email [email protected]
28 | Seaways | August 2019 Read Seaways online at http://www.nautinst.org/seaways
WISTA SEMINAR - AUTONOMOUS SHIPPING
Î A WISTA UK (Women’s International Shipping
& Trading Association) conference held in
London on the Day of the Seafarer examined
the legal and regulatory issues behind
autonomous shipping and pondered the future
of seafarers in a future world featuring remote-
controlled vessels.
The connectivity revolution
Describing automation of vessels as a
‘revolution by evolution’, Filip Koscielecki, a
solicitor and claims executive with the UK P&I
Club, questioned whether international
conventions and national laws were ready even
though the technology was available.
The scale of the technological revolution is
huge, he pointed out. ‘By 2020 we will have 200
billion IOT [internet of things] devices in use
around the world. More and more of them are
on ships, including bilge alarms, ballast water
and performance sensors and cargo monitors.’
From these, the industry could hope to gain
fuel efficiency, smoother container
transhipment and savings in terms of vessel
manpower. Machine learning is already being
used in shipping, especially for predicting
demand and maintenance needs. These
innovations will require huge computing power
on vessels, but ever-wider internet coverage
and cloud computing mean that shipping
companies will be able to buy these services
rather than having to invest in a lot of
equipment.
Mr Koscielecki said the first benefit for
shipping companies would be a reduction in
crew costs, which represent an estimated 30%
of running costs. In addition, he pointed out
that an estimated 75% of accidents are
attributed to human error at a cost to the
industry of $1.6 billion a year. Piracy (apart from
cyber attacks) is likely to become less attractive
if there are no personnel on board, he
suggested. Shipping companies could generate
additional revenue by being able to operate in
less accessible and dangerous areas, and also
being able to use for revenue-earning cargo the
space currently needed for living
accommodation.
Ship construction costs may reduce as more
standardisation becomes possible, although Mr
Koscielecki acknowledged there would possibly
be other costs. Maintenance costs could
diminish with simpler designs; bunker costs
could be cut as different fuels could be utilised
and vessels went slower, leading to easier
compliance with emission regulations.
Legal issues
Legal controls present a more complicated
picture, as there are several issues preventing
an overnight revolution in autonomy. When
approaching legislation, the IMO considered
four levels of automation for its classification,
the European Union three and classification
societies more.
‘All these different levels are difficult to
transpose. There is no single definition of
autonomy and we need to have one,’ he
observed. Although international law seems to
accommodate autonomy Mr Koscielecki made
the point that when law-makers were drafting
conventions they did not consider that ships
would operate without people aboard.
Turning to national laws, Koscielecki said
some definitions at first sight looked broad. For
instance, in the UK the Merchant Shipping Act
1995 states: ‘ship includes every description of
vessel used in navigation’. However, in France
the definition is of manned vessels. In this
context he considers port state control to pose
the biggest problem to autonomous vessels.
‘There are jurisdictional problems,’ he said. Ships
are movable assets so they can go into several
different jurisdictions during one voyage, they
are crewed and managed in other jurisdictions
and have international crews. They could be
controlled from yet another jurisdiction – or, if it
is considered there is no human control, is the
question of where the software was domiciled
relevant?Role of the Master
There are potential legal problems if vessels had
no Master or crew. Masters are a focal point- they are obligated and liable in civil and
criminal law. There will still have to be ‘Masters’
even if they are controlling vessels from land.
Masters act on behalf of vessels with local
authorities, port state control and cargo
interests and ensure seaworthiness. Koscielecki
is particularly concerned about how cargo will
be monitored in terms of port state control, as it
is the Master who must produce documents.
‘How will that be managed?’ he asked.
Another problem with the absence of Masters
concerns the need to render assistance. ‘This is
important, as shipping is a dangerous industry.
But conventions don’t say that it’s people that
have to give assistance,’ he pointed out.
‘Technology could help; for instance, drones
could drop buoyancy aids to people in distress.’
As obligations are increasingly delegated to
shore operators, he foresaw a new maritime role
that would potentially replace that of Masters
and crews. ‘This could go two ways,’ he said,
‘with a direct connection to owners which is
closer to the traditional role or delegated to an
independent contractor.’ These independents
would probably be closer to the technology
vendor, possibly a corporate or individual. Either
way, they will be expected to have some
liability. In establishing that, the contractual
relationship with the owner will be key.
Turning to the qualifications demanded of
those in control of ships, Koscielecki said there
was clear conflict between existing rules and
autonomous vessel operation and this was an
area that needed regulatory change. STCW, for
instance, requires that ‘at no time shall the
bridge be left unattended’ and the OOW is
required to ‘keep the watch on the bridge.’
All of this is pertinent when it comes to
liability, as most jurisdictions are fault-based.
National regulations present a mosaic of
differing approaches, and international
regulation needs to be more uniform. It would
be challenging to define fault under existing
rules should humans not be involved. This could
revolve around software updates, maintenance
and even the choice of vendor.
In certain jusrisdictions, there is a risk that
victims will not be able to claim compensation
if fault cannot be established. If, under such
fault-finding conventions, fault were to be
found in an autonomous vessel, Koscielecki
asked if this could be extended to cover
malfunction caused by software.
In a parallel to the obligation to ensure
seaworthiness, he thought that there could be a
need for software designers to meet certain
obligations. In terms of cyber security, for
example, he commented: ‘I can’t emphasise
enough that ships will have to be prepared to
withstand attacks and if they can’t then they will
be deemed not to be seaworthy.’
Bridget Hogan
Director of Publications