Autonomy: Risk Assessment 24.6 Legal Risks 541
Part B | 24.6
emerging during this time could be discovered while
the vehicle was in open water, and recovery effected.
The appropriate monitoring period is determined from
the initial reduction in probability of survival in the
Kaplan–Meier plot (Fig. 24.6).
More recently, the same approach was used for estimating the probability of losing an ISE Explorer vehicle
in the Arctic. The risk profile for the ISE Explorer under sea ice is presented in Fig. 24.6. Post-campaign
a risk assessment exercise was carried out to calculate
the observed risk. This profile is presented in full line
in Fig. 24.6. Results show that following the mitigation,
the maximum error between the predicted risk and the
observed risk was 9% [24.14].
24.6 Legal Risks
The international rules for the law of the sea, as embedded in conventions such as the UN Convention on
the Law of the Sea (UNCLOS) and the Convention on
the International Regulations for Preventing Collisions
at Sea (COLREGS) are clear with regard to the obligations of ships or vessels. These legal instruments were
written and agreed before the use of AOVs became
widespread in the academic, military and commercial
arenas. It is not surprising therefore that it is not clear
at all whether, in law, AOVs are indeed ships or vessels. The situation is further muddied by the possibility
for a vehicle to be a ship for the purposes of one law,
while not being a ship for another law. Consequently,
the question, Is an autonomous ocean vehicle a ship?
needs to be asked for each international Convention or
national law. Only after the question has been answered
each time can the legal issues, and the legal risks, within
the scope of that Convention or law be fully understood.
Unfortunately, as made clear in the exhaustive analysis
in Brown and Gaskell [24.45] there is rarely a definitive
answer to this question.
Given this situation, the vehicle owner or operator
may face legal risks following incidents if their interpretation of laws or Conventions is found to be at variance
with the opinions of courts. For example, if a court took
a view that AOVs were vessels for the purposes of the
Standards of Training, Certification and Watchkeeping
(STCW) Convention 1995 their autonomous use on the
surface would probably be proscribed [24.45]. It is not
surprising therefore that action is being taken to prevent wholesale blocks on the use of unmanned systems
in the ocean. These actions may be legal, procedural, or
technical.
First, examining the legal context, James Kraska,
a lawyer with the US Judge Advocate General’s office,
considers that the way forward is to recognize AOVs
as vessels [24.46]. His opinion is that the liberal legal architecture of the key international conventions
would allow for inclusion of autonomous systems. If he
were proved correct then they would indeed gain benefit from being considered vessels. For example, the need
to register an autonomous ocean vehicle as a vessel
would provide for state protection. Rights in international law are given to international persons, such as
states, and not to individual artifacts, such as an AOV.
But a vehicle could benefit from those rights if it was
clearly linked to a state, and hence the usual practice
of registry under a particular state and flag. Henderson
reinforced this point for unmanned underwater vehicles
(UUV) [24.47]:
It is in the interest of the United States to establish
the sovereignty of its UUVs to protect them from foreign seizure.
Bork et al. [24.48] agree with Kraska that there is
a need for action, and that a proper legal regime would
give protection to autonomous (and free drifting) ocean
vehicles, but they are not as sanguine that the existing
framework of international maritime law can provide
what is needed. They call for a new legal regime, but it
is not probable that one will soon come about [24.48].
One aspect of international law that is being followed by many is that of seeking permits for use of
AOVs in the exclusive economic zones of foreign nations, in accordance with UNCLOS Part XIII. Showalter gives an accessible summary of these issues [24.49].
Notable precedents are being set particularly relevant
to the command and control of these vehicles. For example, the UK Foreign and Commonwealth Office has
accepted that it is where the vehicle is piloted which determines which nation applies for diplomatic clearance
from the foreign nation, not the nationality of the vessel
making the deployment. Of course, had the vehicle the
benefit of registration and a flag state in its own right,
this question would be moot, but at least this acceptance
is coherent with the outcome if indeed the vehicle did
have a flag state.
Given strict interpretations of international maritime conventions, where autonomous vehicles would
be considered as vessels, they would be hard to adhere to, consequently, many organizations and users
are preferring self-regulation through collective codes
of practice. An early attempt was the code of practice for AUVs produced by the AUV legal working
group of the Society for Underwater Technology in
2000 [24.50]. With representation from military, indus-
Part B | 24.6
emerging during this time could be discovered while
the vehicle was in open water, and recovery effected.
The appropriate monitoring period is determined from
the initial reduction in probability of survival in the
Kaplan–Meier plot (Fig. 24.6).
More recently, the same approach was used for estimating the probability of losing an ISE Explorer vehicle
in the Arctic. The risk profile for the ISE Explorer under sea ice is presented in Fig. 24.6. Post-campaign
a risk assessment exercise was carried out to calculate
the observed risk. This profile is presented in full line
in Fig. 24.6. Results show that following the mitigation,
the maximum error between the predicted risk and the
observed risk was 9% [24.14].
24.6 Legal Risks
The international rules for the law of the sea, as embedded in conventions such as the UN Convention on
the Law of the Sea (UNCLOS) and the Convention on
the International Regulations for Preventing Collisions
at Sea (COLREGS) are clear with regard to the obligations of ships or vessels. These legal instruments were
written and agreed before the use of AOVs became
widespread in the academic, military and commercial
arenas. It is not surprising therefore that it is not clear
at all whether, in law, AOVs are indeed ships or vessels. The situation is further muddied by the possibility
for a vehicle to be a ship for the purposes of one law,
while not being a ship for another law. Consequently,
the question, Is an autonomous ocean vehicle a ship?
needs to be asked for each international Convention or
national law. Only after the question has been answered
each time can the legal issues, and the legal risks, within
the scope of that Convention or law be fully understood.
Unfortunately, as made clear in the exhaustive analysis
in Brown and Gaskell [24.45] there is rarely a definitive
answer to this question.
Given this situation, the vehicle owner or operator
may face legal risks following incidents if their interpretation of laws or Conventions is found to be at variance
with the opinions of courts. For example, if a court took
a view that AOVs were vessels for the purposes of the
Standards of Training, Certification and Watchkeeping
(STCW) Convention 1995 their autonomous use on the
surface would probably be proscribed [24.45]. It is not
surprising therefore that action is being taken to prevent wholesale blocks on the use of unmanned systems
in the ocean. These actions may be legal, procedural, or
technical.
First, examining the legal context, James Kraska,
a lawyer with the US Judge Advocate General’s office,
considers that the way forward is to recognize AOVs
as vessels [24.46]. His opinion is that the liberal legal architecture of the key international conventions
would allow for inclusion of autonomous systems. If he
were proved correct then they would indeed gain benefit from being considered vessels. For example, the need
to register an autonomous ocean vehicle as a vessel
would provide for state protection. Rights in international law are given to international persons, such as
states, and not to individual artifacts, such as an AOV.
But a vehicle could benefit from those rights if it was
clearly linked to a state, and hence the usual practice
of registry under a particular state and flag. Henderson
reinforced this point for unmanned underwater vehicles
(UUV) [24.47]:
It is in the interest of the United States to establish
the sovereignty of its UUVs to protect them from foreign seizure.
Bork et al. [24.48] agree with Kraska that there is
a need for action, and that a proper legal regime would
give protection to autonomous (and free drifting) ocean
vehicles, but they are not as sanguine that the existing
framework of international maritime law can provide
what is needed. They call for a new legal regime, but it
is not probable that one will soon come about [24.48].
One aspect of international law that is being followed by many is that of seeking permits for use of
AOVs in the exclusive economic zones of foreign nations, in accordance with UNCLOS Part XIII. Showalter gives an accessible summary of these issues [24.49].
Notable precedents are being set particularly relevant
to the command and control of these vehicles. For example, the UK Foreign and Commonwealth Office has
accepted that it is where the vehicle is piloted which determines which nation applies for diplomatic clearance
from the foreign nation, not the nationality of the vessel
making the deployment. Of course, had the vehicle the
benefit of registration and a flag state in its own right,
this question would be moot, but at least this acceptance
is coherent with the outcome if indeed the vehicle did
have a flag state.
Given strict interpretations of international maritime conventions, where autonomous vehicles would
be considered as vessels, they would be hard to adhere to, consequently, many organizations and users
are preferring self-regulation through collective codes
of practice. An early attempt was the code of practice for AUVs produced by the AUV legal working
group of the Society for Underwater Technology in
2000 [24.50]. With representation from military, indus-