198
Sensemaking in Safety Critical and Complex Situations
flights. The dominant failures were in power systems, ground control system and navigation systems.
The risks of UAS operations are dependent on the operational domain, i.e. the
type of operation ( delivery, data collection, surveillance, inspection photography,
etc.) and physical details of the drone such as weight, speed and height of operation.
EASA ( 2016) has estimated the probability of fatality of different UAS weights and
estimated probability of fatality as 1% with a UAS weight of 250 g, but 50% fatality
with a weight of 600 g in case of a collision with a human when the drone drops.
Examples of undesired incidents from UAS are: collisions with personnel; interference with infrastructure ( infrastructure such as airports is vulnerable and interference may lead to disruption of air traffic); actual damage to critical infastructure;
damage to the drone; using the drone to spy or steal data (leading to loss of privacy,
data theft and possible emotional consequences). Automated systems and UAS are
vulnerable to attacks through the cyber-physical systems it consists of, such as sensors, actuators, communication links and ground control systems. As an example,
an Iranian cyber warfare unit was able to land a US UAS based on a spoofing attack
modifying the GPS data ( Altawy et al., 2017).
There are several challenges of UAS operations in challenging climatic conditions
such as low temperature, wind, winter with sleet and snow. Operational equipment
may not be tested or hardened for these demanding conditions; thus, requirements,
testing and certification are needed. Communication infrastructure is also demanding in the north, from 70° the quality of satellite communication is degraded. GPS
spoofing may be a challenge and must be mitigated.
Lessons Learned That May Be Transferred
Automation in aviation has succeeded in establishing a high level of safety, due to
systematically automating simple tasks and reducing demands on the pilot: base
development on the science of human factors, building infrastructure, to control
and support flights, strong focus on learning from small incidents and accidents
and support from control centres that have strict control of the operational domain/
operational envelope. Thus, systematic development and stepwise refinement has had
a huge success in terms of safety and trust, in addition to the strong focus on keeping
the human in the loop supported by sensemaking. Even in this environment of high
reliability, there is a strong need to ensure compliance with human factors design
standards and support for human factors assessment in aircraft testing and certification to avoid fatalities by automation as seen in the Boeing 737 Max accidents.
The reliability of drones is lower than for manned planes, and there is a need to
develop improved reliability of the new technology. Systematic risk assessment is
needed to mitigate the areas with the most risks. The HMI between automation and the
human operator is challenging. Design must use best human factors practices to support
sensemaking and ensure that the operator can intervene and take control when needed.
autonomy in rail
By automated metros ( rail systems), we mean systems where there is no driver in
the front cabin, nor accompanying staff, also called Unattended Train Operation
Sensemaking in Safety Critical and Complex Situations
flights. The dominant failures were in power systems, ground control system and navigation systems.
The risks of UAS operations are dependent on the operational domain, i.e. the
type of operation ( delivery, data collection, surveillance, inspection photography,
etc.) and physical details of the drone such as weight, speed and height of operation.
EASA ( 2016) has estimated the probability of fatality of different UAS weights and
estimated probability of fatality as 1% with a UAS weight of 250 g, but 50% fatality
with a weight of 600 g in case of a collision with a human when the drone drops.
Examples of undesired incidents from UAS are: collisions with personnel; interference with infrastructure ( infrastructure such as airports is vulnerable and interference may lead to disruption of air traffic); actual damage to critical infastructure;
damage to the drone; using the drone to spy or steal data (leading to loss of privacy,
data theft and possible emotional consequences). Automated systems and UAS are
vulnerable to attacks through the cyber-physical systems it consists of, such as sensors, actuators, communication links and ground control systems. As an example,
an Iranian cyber warfare unit was able to land a US UAS based on a spoofing attack
modifying the GPS data ( Altawy et al., 2017).
There are several challenges of UAS operations in challenging climatic conditions
such as low temperature, wind, winter with sleet and snow. Operational equipment
may not be tested or hardened for these demanding conditions; thus, requirements,
testing and certification are needed. Communication infrastructure is also demanding in the north, from 70° the quality of satellite communication is degraded. GPS
spoofing may be a challenge and must be mitigated.
Lessons Learned That May Be Transferred
Automation in aviation has succeeded in establishing a high level of safety, due to
systematically automating simple tasks and reducing demands on the pilot: base
development on the science of human factors, building infrastructure, to control
and support flights, strong focus on learning from small incidents and accidents
and support from control centres that have strict control of the operational domain/
operational envelope. Thus, systematic development and stepwise refinement has had
a huge success in terms of safety and trust, in addition to the strong focus on keeping
the human in the loop supported by sensemaking. Even in this environment of high
reliability, there is a strong need to ensure compliance with human factors design
standards and support for human factors assessment in aircraft testing and certification to avoid fatalities by automation as seen in the Boeing 737 Max accidents.
The reliability of drones is lower than for manned planes, and there is a need to
develop improved reliability of the new technology. Systematic risk assessment is
needed to mitigate the areas with the most risks. The HMI between automation and the
human operator is challenging. Design must use best human factors practices to support
sensemaking and ensure that the operator can intervene and take control when needed.
autonomy in rail
By automated metros ( rail systems), we mean systems where there is no driver in
the front cabin, nor accompanying staff, also called Unattended Train Operation