194
Sensemaking in Safety Critical and Complex Situations
threats, technological changes ( with increased complexity and couplings), poor sensemaking, lower possibility for meaningful human control ( Human not in the loop)
and limited learning from accidents.
The term “ Human in the loop” means that the human is a part of the control loop,
i.e. that the human receives information and can influence other parts of the chain of
events ( Horowitz and Scharre, 2015). A key issue is the ability of the actors to make
sense of the situation. In our study, we define sensemaking in a pragmatic context as
a continuous process of interpreting cues to establish situational awareness in a social
context, as described in Kilskar et al. ( 2020).
When trying to scope risks of autonomous systems, we must include regulation, risk governance, organisational framework, interfaces between humans and
the autonomous system, and the available infrastructure ( software components and
cyber-physical systems) to build a sense of the situation for humans and the automated system ( Johnsen et al., 2019).
Autonomous systems are socio-technological systems. Hence, a holistic approach
is necessary, rather than a reductionist approach looking at the system as isolated
processes and components. We lack statistical evidence for the probability of accidents with autonomous transportation systems. However, several actors have started
pilots with different levels of autonomy within different transport modes. There is a
need to collect and systemise experiences from these. The following sections present
a review of experiences from different transport modes. The main objective has been
to gather experiences and status on different transport domains and to learn between
the modes, by asking the following research questions:
1. What are the major safety and security challenges of autonomous industrial
transport systems?
2. What can the various transport modes learn from each other regarding
safety and security related to sensemaking and meaningful human control?
3. What are the suggested key measures related to organisational, technical
and human issues?
FINDINGS
autonomy at Sea
Several countries have developed test areas for testing Maritime Autonomous Surface
Ships ( MASS). The International Maritime Organisation ( IMO) currently uses the
term MASS for any vessel that falls under provisions of IMO instruments and which
exhibits a level of automation that is currently not recognised under existing instruments. There are already several small‐size unmanned and autonomous maritime
crafts which have been engaged in surface navigation, scientific activities, underwater operations and specific military activities.
In Norway, three national testing areas have been established, with supporting
infrastructure, with the aim to test out MASS in the same area as conventional ships.
Norwegian Forum for Autonomous Ships ( NFAS, 2020) is a network established
for sharing experiences and research within the subject of autonomous ships, with
Sensemaking in Safety Critical and Complex Situations
threats, technological changes ( with increased complexity and couplings), poor sensemaking, lower possibility for meaningful human control ( Human not in the loop)
and limited learning from accidents.
The term “ Human in the loop” means that the human is a part of the control loop,
i.e. that the human receives information and can influence other parts of the chain of
events ( Horowitz and Scharre, 2015). A key issue is the ability of the actors to make
sense of the situation. In our study, we define sensemaking in a pragmatic context as
a continuous process of interpreting cues to establish situational awareness in a social
context, as described in Kilskar et al. ( 2020).
When trying to scope risks of autonomous systems, we must include regulation, risk governance, organisational framework, interfaces between humans and
the autonomous system, and the available infrastructure ( software components and
cyber-physical systems) to build a sense of the situation for humans and the automated system ( Johnsen et al., 2019).
Autonomous systems are socio-technological systems. Hence, a holistic approach
is necessary, rather than a reductionist approach looking at the system as isolated
processes and components. We lack statistical evidence for the probability of accidents with autonomous transportation systems. However, several actors have started
pilots with different levels of autonomy within different transport modes. There is a
need to collect and systemise experiences from these. The following sections present
a review of experiences from different transport modes. The main objective has been
to gather experiences and status on different transport domains and to learn between
the modes, by asking the following research questions:
1. What are the major safety and security challenges of autonomous industrial
transport systems?
2. What can the various transport modes learn from each other regarding
safety and security related to sensemaking and meaningful human control?
3. What are the suggested key measures related to organisational, technical
and human issues?
FINDINGS
autonomy at Sea
Several countries have developed test areas for testing Maritime Autonomous Surface
Ships ( MASS). The International Maritime Organisation ( IMO) currently uses the
term MASS for any vessel that falls under provisions of IMO instruments and which
exhibits a level of automation that is currently not recognised under existing instruments. There are already several small‐size unmanned and autonomous maritime
crafts which have been engaged in surface navigation, scientific activities, underwater operations and specific military activities.
In Norway, three national testing areas have been established, with supporting
infrastructure, with the aim to test out MASS in the same area as conventional ships.
Norwegian Forum for Autonomous Ships ( NFAS, 2020) is a network established
for sharing experiences and research within the subject of autonomous ships, with