To drone, or not to drone : A qualitative study in how and when information from UxV should be distributed in rescue missions at sea

(1)

Linköping university | Department of Computer Science Master thesis 30HP | Cognitive Science Spring 2020|LIU-IDA/KOGVET-A--20/013--SE

Linköping University SE-581 83 Linköping, Sweden +46 013 28 10 00, www.liu.se

To drone, or not to drone

A qualitative study in how and when information

from UxV should be distributed in rescue missions

at sea

Author: Rickard Laine

Supervisors: Björn Johansson & Peter Berggren Examinator: Arne Jönsson

(2)

(3)

Copyright

The publishers will keep this document online on the Internet – or its possible replacement – for a period of 25 years starting from the date of publication barring exceptional

circumstances.

The online availability of the document implies permanent permission for anyone to read, to download, or to print out single copies for his/hers own use and to use it unchanged for noncommercial research and educational purpose. Subsequent transfers of copyright cannot revoke this permission. All other uses of the document are conditional upon the consent of the copyright owner. The publisher has taken technical and administrative measures to assure authenticity, security and accessibility.

According to intellectual property law the author has the right to be mentioned when his/her work is accessed as described above and to be protected against infringement.

For additional information about the Linköping University Electronic Press and its procedures for publication and for assurance of document integrity, please refer to its www home page: http://www.ep.liu.se/

(4)

(5)

Abstract

Swedish maritime rescue consists of a number of resources from various organizations that will work together and achieve a common goal, to save people in need. It turns out that information is a significant factor in maritime rescue missions. Whether you are rescuer at the accident scene or coordinating the rescue mission from the control center, information

provides you better situation awareness and knowledge of the situation, which creates better conditions in order achieve the goal for the mission.

Applying Unmanned Vehicles (UxV) for Swedish maritime rescue means another resource that can provide additional necessary information. In this study, several methods have been used to find out where in the mission information from UxVs can conceivably contribute. The study identifies three critical situations where there is a need for UxV. This result, in turn, leads to other questions, such as who should be the recipient of the new information and how it affects the information flow as a whole? Information visualization proves to be an important factor in this. Where you can help the recipient of the information in their work with the help of clear and easily understood visualization without affecting the flow or coordination in their work.

(6)

(7)

Acknowledge

First and foremost, I would like to thank Combitech who introduced me to the WARA-PS project which made the thesis possible. I would also like to thank all the people at Combitech who somehow helped me during the semester, especially my supervisor Emma Jonsson. I would also like to give a big thank you to Oscar Lundblad who has been both literally and figuratively my co-driver throughout this work. Further I want to thank my supervisors Peter Berggren and Björn Johansson for great discussions and suggestions regarding the subject. Also, a thank you to all the participants from all the organisations that has been so kind participating in the study. Finally I want to thank my family who has been a momentous support for me.

(8)

(9)

5.2.1 SSRS (participant 11) ... 29 5.2.2 JRCC (participant 12) ... 29 5.3 Summary of results ... 30 6 Discussion ... 32 6.1 Research question 1 ... 32 6.2 Research question 2 ... 33 6.3 Research question 3 ... 36 6.4 Research question 4 ... 37 6.5 Method discussion ... 39 7 Conclusion ... 42 8 References ... 43 9 Appendix ... 47

(11)

(12)

1

1 Introduction

Search and rescue operations is a fight against the clock. The rescue staff is expected to make quick but at the same time thoughtful decisions. In all forms of rescue missions it is the search and rescue of people in danger that is the most vital. This also includes sea rescue missions (De Cubber, Doroftei, Serrano, Chintamani, Sabino & Ourevitch, 2013).

Research on whether autonomous vehicles can be used in rescue missions has increased significantly and much indicates that autonomous vehicles can provide good support during an ongoing mission. (Kurowski, Martin, & Bernhard P Lampe, 2014; Yeong, 2015)

A rescue mission is about making quick decisions, there is no time to analyse options for too long. Automation is valuable in these situations in that there are abilities to collect and analyse data significantly faster than humans. In situations where time does not play a big role, assistance from automation do not necessarily the same advantageous, since people have the opportunity to take the time needed to analyse manually to make a good decision.

(Johnson, Kuchar & Oman, 2002)

Unmanned vehicles (UxV) refers to all types of vehicles that do not require a human pilot on board. An UxV can be remote controlled by a human or it can be automatically driven with assigned missions. In search and rescue operations, UAVs have already shown their value by helping respondents target their search efforts while avoiding hazards (Doherty & Rudol, 2007). In 2006, two UAVs were used to map the devastated region throughout search of stranded survivors in the wake of Hurricane Katrina. (NSF, 2005) The awareness of the feasibility of using UAVs for search and rescue purposes is demonstrated by a growing array of works for accident identification, route preparation and mission allocation in the fields of image recognition. (Doherty & Rudol, 2007; Goodrich, Morse, Gerhardt, Cooper, Quigley, Adams & Humphrey, 2008)

The Swedish Maritime Rescue is based on collaboration between many organizations where the State, the Municipality and associated organizations at sea work together with the mission to save lives. A visual organisational overview has been created, which can be found in appendix A. Collaboration at sea is a prerequisite for an effective rescue work. The Swedish Maritime Administration handles maritime as well as air rescue within Swedish territories. Joint rescue (JRCC) is a unit within the Maritime Administration that manages and

coordinates maritime rescues 24 hours a day, all year around. The Swedish Maritime

Administration itself has resources to assist in many different rescue operations, such as pilot boats and helicopters, but in reality there are resources from other organisations that are the greatest assets during an operation. These organisations include, for example, the SSRS, the Emergency services, the Coast Guard, the Police as well as the Swedish Military. There is also

a law that determine that private boat owners must help in case of emergency at sea. JRCC is the

coordinator and operator for the rescue missions, but sometimes when the scope of the mission is large and many resources from different organizations are involved JRCC needs coordination aid from the actual scene. That is when the On scene coordinator (OSC) comes in play. The OSC is determined by JRCC and usually it is the first boat on scene or the first one with enough experience that gets that role.

(13)

2

In the event of an accident, SOS alarm will be the first to receive the call. SOS Alarm gather as much information as possible of the situation, using the information obtained through the alarm as well as through supplementary questions. When they acknowledge that the accident takes place at sea, relevant resources will be connected into the case. JRCC will then overtake control over the call from SOS Alarm and become the coordinator of the mission. During the call, when information is being obtained from the caller, the relevant resources are informed of the accident and are already on their way.

Sweden is a country with plenty of water around it. The above mentioned organisations try to cover as much of these water areas as possible, but there is an uneven coverage of resources. In the end, it is those closest to the accident that help. This contributes to the fact that once an accident occurs, the involvement of exactly which resource or individual specialist may differ significantly depending on availability at that particular location. When an accident occurs on water, this means that there can be a many different constellations of resources working together at sea, but still coordinated remotely by the JRCC. A brief explanation of how the process goes during a rescue mission can be seen in Appendix B.

Everyone involved in a specific rescue operation becomes a team with a joint goal. Team and teamwork is a well-researched area and there is one particular factor that is recurring as an important part of a functioning teamwork and that factor is communication. (Hackman 1987; Wilson, Salas, Priest, & Andrews 2007; Fussell, Kraut, Lerch, Scherlis, McNally, & Cadiz, 1998) Information is very important during a rescue operation. With information you have prerequisite for gaining situational awareness (SA) (Endsley, 1995). Through SA a rescuer can get a quicker look at the situation and determine what actions to prioritize. One specific individual cannot do everything by himself, therefore it is important that the information is distributed to everyone as fast as possible. This is done through good communication between the parties. Another important element that is based on good communication is coordination. During a rescue operation, we have JRCC, who is monitoring the operation and communicate with OSC in place. OSC communicate to all relevant resources for the accident. Coordination is centralized in this case. The flow of information between everyone involved goes back and forth and it very is important that this is coordinated properly in order for the rescue mission to succeed.

Applying UxV during rescue operations at sea would mean to add at least one more resource to the team, which also means an additional source of information. The more resources involved, the greater the volume of information, which has been shown to lead to information overload (Fullers et al., 1998). This domain is lacking research on what type of information that actually is desired or needed from the rescue personnel's perspective. It is also important to explore how the information from UxV has an impact on the information flow. Can this result in information overload or can it reduce the information flow, by replacing information previously shared with the rescue personnel?

Today, communication between individuals during a rescue operation takes place primarily verbally via the communication system Rakel and the radio VHF. UxV currently does not have the ability to communicate with the person verbally. The communication with UxV is primarly visual, by, for example, displaying images from the cameras that are set on UxV. The visual information must be very clear and concrete in order for the receptors to understand it and to work effectively. There is no time, for example, for JRCC to sit and

(14)

3

watch live stream from UxV's cameras. This work must be done through information visualization. Images or illustrations have strong advantages in many types of information, and in many cases the combination of both images and text often has the best desired effect (Ware, 2020). Effective information design can accomplish a lot and the main objective should be to reduce difficulties in interpreting information. This is to avoid questions such as what type of information it is, what the purpose is and why it is something the reader should note. Due to the application of UxV within sea missions, there is a gap in the research on how the information from UxV should be visualized for the receivers of the information.

1.1 Purpose

We are soon approaching a time where we will start using UxV in SAR missions at sea. There is a lot of research on the technical part, but there is no research on how UxV and rescue personnel should be able to work together. Rescue missions are based on good

communication, where information can be shared and teams coordinated in the best possible way. Adding UxV, would mean to add an additional resource that can communicate with the rescue team. This study aims to fill part of the knowledge gap that exists today regarding the collaboration between UxV and rescuer, by first and foremost mapping out how

communication between all resources looks today. In order to find out how the new

information from UxV will affect the flow of information, it is essential to explore what type of information the rescue personnel want from UxV and how this should be visualized.

1.2 Research questions

• In a scenario with a drifting boat on state water, what are the critical situations where information from UxVs can be used?

• Who is intended to be the recipient of information and how should the information be coordinated?

• How does the flow of communication change when applying UxV? • How could the information from the UxV be visualised?

1.3 Delimitations

This thesis will only focus on rescue missions at sea and be primarily about the information flow that is being distributed all around the included actors during an operation. There are no such thing as a typical accident at sea. An operation by the SAR team during an accident can look very different regarding on many reasons. However, in order to get this thesis less theoretical and more practical a real case scenario has to be made. This case will be based on data from real accidents that has occurred. All the results will thereafter be based on that scenario. This does not mean that the results will not be adaptable on other scenarios.

One of the research questions will be regarding the presentation of the information. The thesis will go in to the area of information design, but no actual design will be made because the scope of the thesis cannot cover that. It will instead result in suggestions for how the information can be presented based on the data and research in information design.

At present time there are rules that prohibit unmanned vehicles to operate autonomously by themselves in the open nature. These rules will be ignored for the sake of the thesis.

(15)

(16)

5

2 Background

2.1 Stakeholders and research program

Combitech are the stakeholders for this thesis. Combitech is an independent consultant company and are part of the defence and security group Saab AB. They are based in

Scandinavia with over 1.900 consultants deployed in Sweden, Finland, Denmark and Norway. Combitech have international customers all over the world. The company are part of the WASP, which is Sweden´s largest individual research program. WASP ambitions is to become one of the world leaders in the areas of autonomous systems, artificial intelligence and software development. WASP is doing this by combining experts in the areas both from the industry and the academical world.

2.2 UxV

UxV is an umbrella term for all types of vehicles that operate without a pilot. UxV can mean vehicles that are driven on land, in the air, on water and even under water. In the WARA-PS project, In the WARA-PS project there are various models of UxV being researched in this now. This study has chosen to focus on two different models. Both models are drones that can fly and the reason why they have been chosen is because they have come a bit in the

development. It is also easier to relate to drones that have become increasingly common in our everyday lives. It is easier for participants to talk about these two UxVs that they can probably relate to more. For an autonomous boat instead, it has not come as far in

development phase, there are also fewer who have seen such a thing in real life, which makes it more difficult to talk about and relate to. These two models will in the thesis mainly be referred to as UxV, although in some cases where only one is discussed, their model names will be mentioned. The two models, the fixed wings and the quadcopter will be explained more below.

The quadcopter (see figure 1) is a multirotor helicopter that is lifted by four rotors. The

quadcopter is equipped with sensors and cameras, which It can record with. It has the ability to avoid objects and can drop materials (First aid kit etc.). The quadcopter has an automatic detection technology which makes it possible to see objects or humans. It is unusable in rain and heavy winds.

Figure 1 The Quadcopter (Property of WASP)

(17)

6

The fixed wing (see figure 2) is another type of drone with no rotors equipped. The idea with

fixed wings is that it should be placed in a number of places on the water and that it can be sent away using a button press, which can be done at a distance. This results in fixed Wings having the ability to get to the scene early and with an equipped camera it has the ability to record the whole session. The fixed wings has a flight time of about 15 min. After 15 minutes it flies into the water. The fixed wings is waterproofed and can float so it can afterwards be picked up from the water.

Figure 2 The fixed wings (Property of WASP)

2.3 Joint Rescue Coordination centre

The Joint Rescue Coordination Centre (JRCC) (See figure 3) are Sweden’s national centre for sea and aircraft guidance. It is located in Gothenburg and share the same building as the Swedish coast guard and the Swedish armed forces centre. JRCC are operating around the clock which means that they have the command of all the incoming calls that regards sea and aircraft rescuing no matter when and where. In case of bigger events or many cases at the same time more people will come in and a stab with designated roles will be created. JRCC are then the ones who coordinate and call in other resources who will be involved in the rescue mission. (Sjofartsverket, 2020a)

(18)

7

Figure 3 JRCC (Property of Tobias Nicander at JRCC)

2.4 On scene coordination (OSC)

In the event of a sea rescue mission that can be considered large, where many different resources are involved and need to be coordinated, an OSC will be appointed to assist JRCC with coordination. This person is appointed by the JRCC and is usually becomes the person who arrives first to the accident site or the person with most experience. Training is required to become an OSC. It is then up to the person asked if it feels urged to take on the role of OSC. OSC's role is to be the link between the JRCC and the rescue personnel on site. All dialogue between them goes via OSC. (IMO, 2006)

2.5 Emergency units at sea

All different types of organizations (which in this thesis will be seen as resources) who have the ability to participate in a rescue operation and are relevant to the thesis will be explained in more detail below.

The Sea Rescue Society is a non-profit association with over 2200 volunteer sea rescuers

who provide sea rescue around the Swedish coast and lakes. When an accident occurs on water, people stationed in the vicinity of the accident are contacted. The rescue personnel are then expected to be on the road after 15 minutes.

(19)

8

The Emergency Service also has the ability to help at sea. All emergency service stations

around Sweden do not have boats. The rescue service belongs to a municipality and also has the task of being involved in all types of rescue efforts, which means that if an accident occurs in the water, not all resources can go on that alarm, there must be resources that can take care of if an emergency occurs on land.

The Coast Guard focus on emergency services, emergency preparedness and maritime

surveillance. When it comes to accidents at sea, they are mainly used for oil spill accidents. They have well-equipped boats for any oil spill accidents. The Coast Guard boats also have well-equipped systems on board, which results in the person appointed to OSC most often using the system on the Coast Guard, regardless of which organization they come from.

2.6 Communication systems at sea

To a certain extent, this work is about communication between the rescue personnel at sea. It is therefore relevant to address what type of communication tools they use.

Rakel is a national digital communication system and is shared by Swedish public safety organisations in order to provide security for the Swedish population. The public safety organisations can communicate safely and effectively both in everyday life and in crisis. These users rely on Rakel for everyday operations and it is still being developed in dialogue with the users. Its owned by the Swedish government but it is being operated and managed by the Swedish Civil Contingencies Agency (MSB) and whenever a new equipment, an

application or a new organization that wants to join the system, MSB are the ones that need to approve it. (MSB, 2020)

Before Rakel, there were over 200 analogue communication systems among the Swedish organizations that worked with public safety. They desired a joint system that would facilitate a cooperation between the organizations. Rakel was deployed 2006 first in the southern parts of Sweden and today Rakel covers 99.84 of the Swedish population. There are a lot of safety and security rules to follow for the organizations when using the system. Sensitive

information as passwords and identities needs to be handle carefully by the users. Rakel uses a European standard mobile radios system called Terrestrial Trunked Radio. The system is built to work even under difficult conditions as power failure or bad weather. Rakel is similar to a regular mobile network, but what differ Rakel from the regular one is the ability to create group communication between the resources. This makes the work for the organizations more effective, more users simultaneously can talk with each other, which make it easier to

distribute the information to the users you want to reach. (MSB, 2020)

Marine VHF radio, also called “The lifeline to shore” is a worldwide system that provides security and constantly open communication with the outside world. VHF-radio can be used to inform people at sea with important messages. People at sea can also use VHF-radio in order to come in contact with other leisure boats and emergency boats. VHF 16 is the main channel for people in need at sea. It provides direct contact with JRCC and other nearby boats which can assist you. (Sjofartsverket, 2020b)

(20)

(21)

10

3 Theory

The theoretical framework will be presented in this chapter. The theory is chosen to give an understanding of the aims of the study and to provide a basis for the reasoning that will be presented during results and discussion. The theory will begin by providing a background in communication and primarily team communication. Communication within teams, in this case rescue teams, play an important role for a well-functioning response to accidents at sea. The theory will then cover SA. Getting SA early is important to get an understanding of the extent of the accident and to know how the effort should be coordinated. For this work, it is mainly interesting to look at SA from the perspective of JRCC and OSC, these two resources are the ones who will coordinate the effort and are in need of information from the rescuer on site in order to obtain a good SA. The theory will then take up coordination and look at what really lies behind good coordination. Rescue mission are currently coordinated in a

centralized way, which will be explored more. This section will then conclude with the theory of information. Information is a broad concept and there is much to look at. For this work, it will be of interest to explore further information flow. When communication takes place and information is to be coordinated effectively, an information flow is created between all involved resources during a rescue operation. The theory part of information will also handle information design. Information design is relevant to this work for the reason that the

information sent from UxV must be visualized in a way so that the information becomes easy and fast to understand.

3.1 Communication

Communication is a well-researched area, although there is not a common definition for what it really means. Griffin (2018, s.6) breaks it down to its simplest form and defines it as

“Communication is the relational process of creating and interpreting messages that elicit a response”. Hackman (1987) studied teams and what parts are needed to accomplish a task effectively and found that the team has to be coordinated in a detailed way. In order for a team to do that it requires team design and communication. Team design can be defined explained as, the way the team is structured and the tools they are using. Communication includes transferring information in every way possible to make it possible for the team to negotiate their goal, make decisions and to share their own status. Research has shown that teams where members can predict each other's need for information are more likely to perform better than teams that exchange information after being asked (Entin et al., 2000; Shah & Breazeal, 2010).

Communication can be used as a general term to include all forms of communication, but here are various terms that can be used to explain in a more precise way how team members

communicate with each other. Wilson et al, (2007) studied teamwork in a battlefield and broke the term into three sub-terms, information exchange, phraseology and closed-loop communication. The former is the most relevant for this study and will be addressed here in greater depth. Information exchange refers to all of the information that is transferred between the sender and the receiver. The adequate information will be passed on to the adequate person who will provide frequent status updates which will result in a better knowledge about the bigger picture. Effective information exchange leads to teams that are better to provide a

(22)

11

clear concordant picture of the situation which will make it easier for the team to build and sustain a situational awareness.

During a SAR at sea, most of the communication between all resources takes place via radio, through either Rakel or VHF radio. This applies mainly to the person at JRCC who leads the work remotely, but also the rescue personnel on site at the accident rarely have time or opportunity to communicate face to face. This places high demands on communication systems to work. Heath and Luff (2000) argue that communication for organizations is

affected by the technology they use. Hutchins (1991) studied a group of soldiers on a ship and found that they communicated according to the environment they were in and the technology they had access to.

3.2 Situation awareness

Situation awareness (SA) can be defined as “the perception of the elements in the

environment within a volume of time and space, the comprehension of their meaning and the projection of their status in the near future” (Endsley, 1995, p. 36). SA is often referred in the context of an operator in a dynamic system. Nowadays systems become more and more complex for every day. This in turn puts more pressure on the operators abilities to act as efficiently as possible. The operators SA of the system becomes a crucial part for them in order to perform well and take good decisions. SA is not solely about being aware of a great number of data regarding your system. SA also requires one to have good understanding of the situation right now but also about the systems future states and the operators intentional goals for the system. (Endsley, 1995)

According to Endsley, (1995) SA shown to be essential in many different domains and

contexts. For example, in tactical and strategic systems which is daily utilized by personnel in rescue services. Here fast decision has to be made and they have to rely on the SA in order to make the decisions. Decisions based on incomplete SA can in these context lead to heavy consequences. In complex and dynamic environments it can become challenging to maintain SA. Time can be critical in complex and dynamic environment and the operator is required to take decisions by quickly analysing the environment. SA will in these cases bring a more understanding of the whole situation and form the basis for their own decision making. According to Endsley (1995) SA can be divided into three hierarchical levels which will be clarified here.

Level 1 SA: Perception of the elements in the environment

The first level regards the operators first perception of the current state. Here the decision maker needs to perceive relevant elements of the environment. Endlsey takes a pilot in a cockpit as an example where the pilot would perceive elements such as other aircrafts, natural objects and varnings signal inside the cockpit.

Level 2 SA: Comprehension of the current situation

In level 2, the decision maker will now comprehend the elements perceived by the operator from level 1 and see how the significant elements can be relevant for the decision makers goal. Here a holistic picture can be created based on the knowledge from the elements in level 1 and other elements.

(23)

12 Level 3 SA: Projection of future status

The highest level of SA is the ability to project the future actions of the elements in the environment. In order to do so the decision maker need knowledge regarding the status and dynamic of the perceived element and the comprehension of the current situation from the former levels.

3.2.1 Shared Situation Awareness

Aiming for SA does not necessary have to do with only one person (Endsley, Bolstad, Jones & Riley, 2003). The authors believe that there are many cases where a team need to attain SA in order to make correct decisions which requires the team to share SA. According to Van de Walle et al., (2016) in an uncertain and complex environment where time pressure is a critical factor, the most essential thing is that a team can obtain SA. Which can in simpler forms be expressed as, having an understanding of the situation they are in. As mentioned before, SA is highly dependent on the information that is acquired, but in order for a team to maintain SA the information that is acquired has to be distributed between the team members. It is also important that the information is easy to understand and that it triggers the team to take action. Seppänen, Mäkelä, Luokkala and Virrantaus (2013) conducted a study in the subject of shared situational awareness in search and rescue mission, where many organizations are involved and must cooperate with each other. The researcher in this study attempt to identify critical factors that affect the formation of shared situation awareness. It turned out to be crucial to identify the information the rescuer needed during the mission and also if the rescuer has gained the understanding of where in the mission information it is important to share your knowledge they will be more motivated to share what they know.

Shared situation awareness between team members enables the team to understand the task they are up to, understanding of what others are doing and further actions that will be

executed (Endsley et al. 2003). It allows team members to predict the information and support needs of other team members, leading to a reduced need for explicit communication with each other and better coordination of actions (MacMillan, Entin, & Serfaty, 2004). For a team to successfully achieve the common goal they must create a mutual understanding of the task they are up against (Mathieu, Heffner, Goodwin, Salas, & Cannon-Bowers, 2000). By exchanging information between a team, a shared SA can be created, but it is important that communication is performed effectively. Results from previous research have shown that good collaboration between teams is about getting information when you need it and not having to ask for it (Crant, 2000; Entin & Entin, 2000; Entin & Serfaty 1999; Shah & Breazeal, 2010)

The definition teams used for this study claims that teams are two or more individuals having a common goal (Dyer, 1984). This goal can both mean that physical and mental strength is needed to solve it. Complex situations are becoming more common in many domains which mean that mental strength is needed. In these cases, it is the cognitive resources that are demanded. Cognition used in a team context refers to cognitive processes or activities that take place at a team level. Teams participate in cognitive activities as a unit, this activity is greater than the knowledge that each team member carries. This is because if an activity is too complex, it becomes almost impossible for an individual to have complete awareness of the situation himself/herself and to be able to see the activity from all multiple perspectives. It is

(24)

13

also unlikely that all team members will have a common perspective. Because of this, the team needs to collaborate and execute the activity together on a team level. (Cooke, Gorman, Myers & Duran, 2013)

3.3 Teams

Teams are according to Dyer (1984) social entities with two or more individuals that share a common goal but have specified roles. The individuals are likely to have different sub-goals than the other team members but are high interdependent by the others sub-goals. Teams are usually organized hierarchically and has to incorporate and share information between each other. Coordination and collaboration are two important components for a team to succeed in its work. One word that can sum up all the important parts of a functioning team is teamwork. Researchers studying teams believe that there should be a discrepancy between group and teams (Duffy 1993; Orasanu and Salas 1993). The argument for this is that members in a team have different tasks and roles which also means that they contribute with information in different stages in the decision process. While people in a group is often fully involved in the whole decision process. Furthermore, teams are more commonly seen as possessing different artefacts and equipment in more complex situations than the generic group.

Teamwork is defined by Salas, Sims and Klein (2004) as a collection of coherent thoughts, behaviour and feeling in each team member that are needed in order for a team to function and to achieve a common goal. Cohen and Levesque (1991) clarifies that it is important with criteria for what can actually be considered teamwork. We would for example not argue that there are teamwork involved in cars driving on a street, even though all of the cars are driving in a coordinated way acting accordingly by the traffic signs and the roads. Yet when a group of drivers coordinate a convoy together and drive to the same destination the group of drivers seems to act as a single agent with the same behaviour, goals and intentions.

3.3.1 Team coordination

Teamwork is not something that just occurs by itself, instead prior work rather tells us that in order for a team to be able to successfully reach their goals they must coordinate their

activities (Cannon- Bowers, Tannenbaum, Salas, & Volpe, 1995; Guzzo & Shea, 1992;

Swezey & Salas, 1992). Without wasting any precious resources team members can sequence, synchronize, integrate and complete tasks through appropriate coordination mechanisms (Cannon-Bowers et al., 1995). Lack of coordination may lead to a certain team member not sharing the common understanding of the situation as the rest of the team, which can lead to serious consequences in a critical situation. Successful teams that coordinates during times of low stress, which ensures that all team members have a shared understanding (Orasanu, 1990).

Stasser and Titus (1985) introduced the hidden profile paradigm, which opened up a new path for the research in information and group decision making. The authors found that group discussion is a poor means to commute new information that is not shared by everyone in that group. Since the discovery of the hidden profile paradigm researcher have tried to reduce the bias by forming experiments where a member in a group gets designated as an expert in a specific domain or by appointing a leader of the group. (Stasser et al., 1995; Larson et al., 1998). During crisis there is need for sharing relevant information, but at the same time it is

(25)

14

crucial to act quickly. The situation can be very dynamic which opens up for an environment where the known information and the information that is needed can change rapidly and incessantly. A combination of correct information and quick information is sought in these conditions.

Brehmer and Svenmarck (1995) explored how geographically distributed teams manage to control a dynamic system. They created two types of teams, where one team's information was central coordinated through hierarchical architecture and the other team was able to communicate with everybody without any hierarchy. They found that the team with central coordination controlled the system better. One reason for the success of the central

coordinated team is that the coordinator in this case had all this information managed to create an image of the entire environment, which the authors calls a second time-scale instead of a constrained local view of the environment. They also found that in the team where there was only one person responsible for coordinating information, there was a risk that the

management of information could become too overwhelming.

It is common for a coordination center to consist of a team that has access to technological tools to be used to assist in their work. Something that is also common is that the team is not physically in place either, but the work is done separately from a completely different position from what they coordinate. This means that they are not the first to receive the information, but they must obtain the information indirectly from resources on site. Which also helps them get no direct feedback. Thus, in order to gain an understanding of the situation, they must rely on the information they can obtain from people on site and artefacts. (Artman & Garbis, 1998)

3.4 Information

What is the point of communication when the information is neither relevant nor adequate? Communication is fundamental but it also relies on that the information that is being communicated is relevant and adequate. A team that works in crisis situations as rescue operations at sea are in need of information that is relevant for that specific situation and also that it is being distributed at the right time. Van de Walle (2016) The International Federation of the Red Cross and Red Crescent Societies explains it clearly as “The right type of

information leads to a deeper understanding of needs and ways to meet those needs. The wrong information can lead to inappropriate, even dangerous interventions.” (IFRC, 2005). According to Salomon, (1993) humans have the ability to interpret true meaning of

information, but the representation of the information will affect how we will interpret it. The author points out that it is special technology that can affect how different information can be represented. Technology can also arrange the information flow through its own architectures which means that the technology that is being used affect and allows the information to be distributed between agents. Choo (2001) explore the information processes in organizations and visualize it through a cycle that explains how the information flow continuously goes through three different modes; sensemaking, knowledge creating and decision making. In this cycle termed, the knowing cycle, the outcome of the information from one mode can create a clarified context for information use in the other modes. You create sensemaking through shared meanings and purpose. While they may not agree about the content of a particular issue, and may adopt diverse positions on how it should be resolved, nevertheless there is collective recognition that these issues are salient to the organization. The outcome of

(26)

15

knowledge creation is new capabilities and innovations that enhance existing competencies or build new ones; generate new products, services, or processes; or extend the range of viable organizational responses to a problem situation. The value of new knowledge is assessed locally by its ability to solve the problem at hand. By structuring choice behaviour through premises, rules and routines, the organization simplifies decision making, codifies and transmits past learning, and proclaims competence and accountability. The last mode in the cycle is decision making. Decision-making is about choosing between different choices that will in turn lead to different situations. Rational decisions are based on having an

understanding of what each decision will have for consequences. This knowledge can be gained by having knowledge and understanding of the situation that is taking place now and their past experiences.

3.4.1 Information visualization

According to Jacobsson (1999), information visualization or information design can be explained as planning and forming a message content, and the context in which it is presented in order to achieve specific objects in relation to user needs. It is crucial to describe something that is difficult to solve or to understand and then explain the solution in a straightforward way. Horn (1999) emphasizes that the discrepancy between information design and other types of design is that information design focuses on impact and efficiency in achieving a communicative purpose. The information you have should be understood as effectively as possible by someone else.

Visualization can be used as a cognitive tool as it an external artifact that man can use, for example in decision making (Ware, 2020). Image or illustration has clear advantages over words when it comes to many types of information. At the same time Ware (2020) points out that a great advantage words have over images or illustrations is that the natural language is ubiquitous. Meaning that the spoken language is the most shared system of symbols. But this also requires that you understand the language that is visualized for you. There is also a design consideration to have in mind about the process of comprehending images. For

example, it may take a long time to scan a complex diagram for all its details. Only a fraction of the information can be extracted at first glance. Studies show that we first understand the overall structure and shapes of an object, and then after a while understand the details (Price & Humphreys, 1989; Venturino & Gagnon, 1992).

There are many studies in this topic that aim to find out what type of medium is best for comprehension. Najjar (1998) found that text is better than images or illustrations for communicating abstract concepts. Chandler and Sweller (1991) studied procedural

information and found that text and spoken languages is the best way to provide that type of information, but also in some cases it is possible to use text together with images. Most researchers agree on the information should be visualized depends on the situation and that a combination of different media that together complement each other is the most effective way of conveying information. (Faraday & Sutcliffe, 1997; Mousavi, Low, & Sweller, 1995; Wadill & McDaniel, 1992)

According to Waern et al. (2015), a combination of image and caption can be seen as the obvious elements where they should be used to reinforce each other. Text is important for the context of the image, as an image can often be ambiguous. Text can have the function to explain the image and to anchor the interpretation that the sender wants the receiver make.

(27)

16

The relationship between text and image must be relevant in order to facilitate or control interpretations of images.

Lipton (2007) state the importance of respecting the content of the information. According to him, visualization should not have a decorative purpose where everything is as visually beautiful as possible, but there must be a meaningful use of the information. This can be done through a combination of graphical design and usefulness, where the forms and shapes can cooperate with the audience and convey content before appearance. Through effective

information visualization, much can be achieved. The aim should be to reduce the difficulties in interpreting the information, thus avoiding questions such as what type of information and what it should be used for. Effective visualization answers these questions by itself.

3.4.2 Ware´s visualization process

Ware (2020) present a design methodology for producing cognitively efficient visualizations. There are seven basic steps that are almost always part of any visualization design process. Ware steps addresses primarily the theoretically design problems and leaves out the practical ones. As an example, it does not explain in what way data should be obtained nor which software tools that can be utilized for the building the visualization. The steps are: 1) a high-level cognitive task description, 2) a data inventory, 3) cognitive task requirements analysis, 4) the identification of visualization types, 5) the identification and choice of cognitively efficient interaction methods, 6) prototyping and application, and 7) evaluation. Steps 6 and 7 address information visualization well into the design process, which this study has no

intention or scope to address. For these reasons, these two steps will only be explained a little briefly, to give the reader an understanding of Ware's entire design methodology. In the rest of the work they will not be written or discussed about.

1. The first step, high-level cognitive task description is about determine the problem that the visualization will solve. At this step no specific details will be discussed, instead a broad goal will be set. Here the anticipated cognitive tools that is going to support the final product will be defined. The cognitive tools can for example be sensemaking, planning or design.

2. The second step revolves around data inventory. Here, the designer should become familiar with the available data. It is fundamental to understand the product that you are about to visualize. It is not only the structure of the data that the designer has to have in consideration for how the data should be visualized, the designer also have to think about the semantic of the data. The data can come in many varied forms and for that reason have different attributes. The following list contains the different

appearance the data can possess.

3. Cognitive task refinement is the third step and as the name reveals, this is where the set of tasks can be refined. Here the overarching goal can be dissected into more analytics questions and from that decide which of those questions that can be answered by visualization. Visualization does not have to be the answer to every problem. A good way to dissect the goal is to apply “Who, What, Where, When, Why, and How” on the problem.

4. Identification of Appropriate Visualization Types is the fourth step and now is the time to think about what type of visualization that is most appropriate to apply. There

(28)

17

are many different types of visualizations to choose between and the author present maps, diagram, charts and composite as the big categories.

5. In the fifth step, applying Visual Thinking Design Patterns for Cognitive Efficiency, the designer can from the chosen visualization type from step four combined with the most effective interactive methods.

6. The sixth step is about prototype development and this is where you go from abstract ideas to more practical ones. Here the designer can sketch their ideas.

7. The final step is evaluation and design refinement. The goal here is to evaluate the final design and see if it meets their task requirements.

3.4.3 Information overload

Information overload coined by Gross (1964) is the overbalance of information that a person can be exposed to when a decision has to be made. A person that is about to make a decision has a limit of cognitive capacity and when the amount of information exceeds that capacity it is likely that the quality of the decision will be affected. (IDF, 2020) Previous studies have shown that when communication increases there is a risk for overload of information

(MacMillan et al., 2004; Garbis & Artman, 2004). MacMillan et al., (2004) points out that it is beneficial to reduce the amount of communication between team members by apply more implicit coordination where team members learn to do their tasks without the coordinator having to give orders.

According to Fussel et al., (1998) when discussing communication we need to understand the trade-offs and interaction between overload and shared mental models. In a time critical situation where team members has to communicate a big amount of information it comes more difficult for the whole team to achieve a shared mental model due to information overload. But when mental models are being shared accurately it can reduce the overload. Applying UxV in rescue operations at sea may soon become standardized. Trujillo et al., (2015) studies this from the operator’s perspective who might have to control several vehicles at the same time. The authors state that it is likely that one operator will control a fleet of autonomous vehicles. The operators need to be able to focus on the overall mission and not on the manual control of the vehicles. It is a risk that an operator in charge of several vehicle can cause information overload. Trujillo et al., (2015) also believes that the operator's control room must be fundamentally changed in order to be applicable.

(29)

(30)

19

4 Method

The methodology for this study will be presented in this section. It contains parts about the participants, the data collection and how all the data have been analysed. The data collection has been performed in several steps. To make it more understandable to the reader and to give an overview, a process model has been created (see figure 3) which visually give an overview of all the steps. These steps are explained in more details in this chapter.

Figure 4 Method process flowchart

4.1 Participants

This study was conducted in many steps and had 12 participants from several different organizations. A table has been created (see table 1) that shows each participant's knowledge background and expertise. The table also shows what type of qualitative method was used for each individual participant. Finally, the table shows what the data was used for, i.e. what type of context it had. More information follow in table 1.

Participants Profession/Expertise Qualitative method Context 1 Lecturer in maritime human

factors at Chalmers

Interview Understanding of automation and maritime

2 Researcher and developer at SSRS

Interview (On the phone)

Understanding of drones and sea rescue

3 Rescue leader for emergency service

Interview Understanding of emergency service work at sea and

(31)

20

the applying of UxVs

4 Rescue leaders for the emergency service

Focus group Creating a scenario and applying of UxVs at sea

5 Rescue leaders for the emergency service

Focus group Creating a scenario and applying of UxVs at sea

6 Rescue leader at SSRS and Swedish Maritime Administration Interview Creating a scenario and applying of UxVs at sea

7 Rescue worker for the emergency service

Interview Needs analysis and applying of UxVs at sea

8 Rescue leader at SSRS Interview Creating a

scenario and applying of UxVs at sea

9 Rescue leader at JRCC Interview Creating a

scenario and applying of UxVs at sea

10 Rescue leader at The Swedish Coastguard

Interview Refining scenario

11 Rescue leader at SSRS Scenario walkthrough Scenario

12 Rescue leader at JRCC Scenario walkthrough Scenario

Table 1: A summary of the participants

4.2 Data collection

The data collection started by interviewing three people with different expertise in order to obtain more knowledge about applying UxVs in rescue missions at sea. Data from these three interviews provided insight on the following three areas:

• Automation in maritime. • UxV in rescue missions.

• Swedish maritime organisations

From the data from these three interviews it was clear that a scenario was suitable to create for this study. This because a scenario would make it less theoretical and instead have a more practical approach. A scenario that can be anchored in users' daily work makes it easier to discuss and thus be able to apply the result from the study in the real world.

(32)

21

To create a scenario, it needed to be based on data on what real rescue missions might look like. To obtain this data, several semi-structured interviews and focus groups were held with several relevant organisations at different locations in Sweden. To develop a scenario, questions were asked about the participants' process when different alarms occur. Some questions were about what type of alarms that are common. There were also questions about the scope of the alarms and what impact different scope has on the process and the involved resources. Obviously, there are significant difference in the process and the resources

involved between a small or a large accident. For example, a small accident could be a person fallen into the water in the local marina, where a big accident could be a larger vessel that has crashed which can lead to fire and oil leakage.

The interviewer also took the opportunity to ask about the participants thoughts on the application of UxVs and where in the rescue mission it could be applicable. Communication and coordination where also subjects that were covered. This data was then used for the thematic analysis. Since rescue mission involve several organisations, it was important that as many resources as possible were interviewed. The different organizations are all well spread throughout Sweden, which also means that their daily work looks different even though they are part of the same organization. A typical accident in one place may not be that common in another part of Sweden. Because of this, it was therefore also important that participants from the same organization but from different places in Sweden were interviewed.

4.2.1 Creating and validating scenario

The data was analysed using a thematic analysis. The scenario was created in a flow chart, where the communication between the resources and the actions performed were mapped. The alarm on which the scenario is based on is a typical recurring alarm that all the resources could relate to. It is a medium-sized alarm involving four different organisations. After the scenario was created, three previous interviewees reviewed the scenario and a few

refinements were made. The scenario is described briefly in text and can be seen below with more detailed information (see figure 4).

SOS Alarm receives an alarm about a drifted boat out in the Västervik archipelago. The caller has seen a boat without passengers. When SOS Alarm understands that the matter is on water (state water), they immediately engage JRCC on hearing and they quickly take over the assignment from the SOS alarm. JRCC is looking for boats that can be found near the drifted boat. They alert to both SSRS Loftahammar and emergency service from Västervik is first at scene and creates an initial report on the situation. JRCC announces that there is a missing person and coordinates a search strategy. SSRS arrives at the site and helps with the search. The person is found on a nearby island.

(33)

22

Figure 4 The final scenario

4.3 Scenario walkthrough

After the thematic analysis was completed, two additional interviews were conducted. One person from JRCC and one person from SSRS. These two interviews were very much based on the scenario and therefore it has been given the name scenario walkthrough. Participants were given access to the scenario while the moderator guided the participant through each step. The participant was told to speak freely about each step and comment on all thoughts that arose. The purpose of this method was to acknowledge how much their thoughts were consistent with what had been identified in the previous analysis. The purpose with this method was to get a confirmation on the findings from the thematic analysis and to discover any new relevant input from the participants.

4.4 Qualitative methods

This section explains the chosen methods used for this thesis; how do they work and why they were chosen.

4.4.1 Interviews

All interviews with the participants were semi-structured. A semi-structured interview is, as the name reveals, a mix between a fully structured interview and an unstructured interview. During a semi-structured interview, the interviewer can come up with pre-written questions while new questions are asked during the interview. Although the interviewer has prepared

(34)

23

some questions to explore the subject, there is no clear structure and flexibility is vital. (Howitt, 2016)

Semi-structured interview is a good fit for this work because it was important that there was a flexibility where the interviewee felt that he could talk about what he thought was interesting about that topic. All interviews were recorded and notes were taken during the sessions. 4.4.2 Focus groups

Data was also obtained using the method, focus group. A focus group means that a group of people gather to discuss a topic for a limited time. There is no direct answer for how many people need to be part of a focus group as long as the focus group is not too large. The word focus stands for the discussion in one way or another touching on a specific topic chosen by the moderator in advance. The moderator's job is to lead the group as often as is required to discuss the chosen subject. The moderator should not be acting as an interviewer, but should facilitate the discussion of the party. Therefore the group will speak openly within the defined framework of the subject. (Wibeck, 2010)

Focus group applies well as a method for this study because it creates a more discussion between the participants unlike the interviews. Especially a discussion where experts in the field can speak freely and guide the discussion more how they want and not have an

interviewer with less knowledge than the people who will direct the conversation. All focus groups were recorded and notes were taken during the sessions.

4.4.3 Ethics

This study followed the four principals provided by Vetenskapsrådet (2002). These four principals are the following: the information principal, consent principal, the confidentiality principal and the utilization principal. All participants had to fill in a consent form before the start of the focus group, interview and scenario walkthrough. The consent form also contained information about the purpose of the study and their rights regarding their participation. See appendix C.

4.5 Thematic analysis

To find common themes from the interviews, a thematic analysis was conducted. According to Bryman (2011), it is one of the most common approaches in the analysis of qualitative data. It is an iterative process where data is divided into central themes and subthemes, after careful reading of the transcription on repeated occasions. The thematic analysis is based on Braun and Clarke's (2006) process which contains a total of six steps. Initially, it is about

acquainting yourself with the collected data to generate the first codes, searching for themes and reviewing them. Finally, define and name the themes that are presented in the report, preferably with elements of narrative representation. The six steps are briefly explained below.

1. Get to know the data. Get yourself familiar with your data. In this case it was the data consisted of transcriptions.

2. Generate initial codes. After going through your data several times, interesting findings will be coded.

(35)

24

3. Search for themes. After coding the data a search for themes begin by finding patterns with codes that repeat itself.

4. Review of themes. Go through the themes and keep the themes that are relevant for the research questions.

5. Define and name themes. Review what each theme cover and put a name that clearly explains the content.

6. Produce the report. When all the themes are defined and named, it is time to explain your themes to the reader in the report.

(36)

(37)

26

5 Result

In this chapter, the results of the data collection will be presented. The result chapter will commence with the findings from the thematic analysis. Four themes have been identified and all of these are important to moving closer to the answers to the research questions. The first three themes covers three critical situation that has been identified in the scenario where information from the UxV have been found to be useful. The fourth theme revolves around ideas that naturally arose from the other three themes and consists of the flow of information and who becomes the recipient of that new information. The second part of the result is based on data from the scenario walkthrough. The chapter then concludes with a summary of all results

5.1 Thematic analysis result

Four themes will be presented below. 5.1.1 Initial phase

Initial SA is an important part in rescue missions. An initial SA will help all the involved personnel to obtain a clear picture of the accident and the scope of it. In many cases when accidents occur it is the person who calls SOS Alarm that will provide all of the information. JRCC will have to coordinate their work from that information. It can be difficult for a person that calls in that have no experience from rescue mission to understand the scope of the accident and to know what type of information to look for that can be key information for JRCC. Taking the scenario as an example, the caller to the SOS alarm have seen a drifting boat on the water and not really give more information than what position it has and what direction it is heading. It is not until the first rescue boat arrives at the scene of the accident that new information can be given.

From the thematic analysis it was clear that this theme identified a problem and that there is a need from the participants to apply UxV in the initial phase of the mission. The JRCC, which coordinates the mission, evidently wants to send out the right amount of resources to the accident site, but it is difficult to know the right amount of resources. JRCC send for that reason rather too much resources than too few. At the same time, one cannot send too many resources to one place, as there must be resources to use if an accident occurs at another location. As mentioned earlier, it is not until the first boat arrives at the scene of the accident that new information can be given. Information that both rescue personnel and JRCC want in this situation, for example, the weather, how many people there have been on the boat, how many can be in the water, if the engine is warm, etc. In the scenario, it is the Emergency Service that arrives first at scene. From their initial understanding of the situation, they determine that this is a drowning accident and that more resources are needed in order to start the search for that missing person. Had they known this earlier, more resources could have been sent out quicker and the search process could have started earlier.

What can be identified from the thematic analysis is that it would be a good application for UxV at this early phase in order to get an initial understanding of the accident. If for example, the fixed wings drone could be dispatched to the accident site as soon as the call from SOS

(38)

27

alarm has come in and they have determined the position. The drone would then be in place much earlier than the first boat and then be able to send relevant information about the accident, and then make it easier for JRCC to know how much resources are needed.

”Although they see weather report and have some idea about how it is blowing, they do not have local idea about how it looks. But once we are in place we do a wind report and that is of utmost

importance so they understand what we need later, in terms of resources.” Participant 5 5.1.2 Overview

Rescue missions are managed, as previously mentioned, by JRCC, who are located in a control room in a completely different place of the country. In its control room, JRCC is equipped with a system for coordinating the mission. They can track where the resources involved are during a mission by following their GPS signal. If the person who is in distress calls in to SOS Alarm, their geographical position is sent directly to the JRCC, which allows them to keep track of where they are. JRCC can also use its system to see if there are nearby resources by the accident that can be called in if more resources are needed. They also get verbal information from the OSC in place. However, as previously mentioned, JRCC is not in place at the time of the accident and does not have a clear overall picture. They must

themselves create a mental picture of the accident site using the information they receive from OSC and the information they can get out of their own system. Rescue personnel on site also see the effort from their own boats and have no opportunity to get a complete overview of the situation

Where UxV could provide the necessary information that is not possible to obtain at current missions. In practice, a UxV could hover over the accident site and literally contribute an overall picture of the situation. Something that came to light when this topic was being discussed by the participants was that there was not really time to look blindly at live stream images during an ongoing mission. JRCC already have their hands full with coordinating the mission, but the option of sometimes being able to watch live stream pictures would

contribute to the mission.

After a rescue mission, each organization writes a report on what have happened. First and foremost because it is public documents that the public should be able to obtain, but also the rescue personnel themselves. Through the documentation they can go through how the rescue mission went, what went well and what could be improved. The report is made by each individual organization that has been involved in the effort and it is clear that there are different levels of how detailed they are. The rescue service turns out to be the organization that is most detailed in its reports and uses the report extensively to learn about the future. A sub-theme that was identified in this theme was the idea of utilize the UxV for

documentation. By having a UxV hovering over the situation and capture the whole process the data could be recorded and used for later debriefing by the organisations. These type debriefing is something that the rescuer already doing in order to review what went good and what can be improved until next time. The participants felt that this type of material would be of great benefit to them.

(39)

28

"Drones can go up and give us an overall picture"

Participant 7 “Documentating the mission is important afterwards. Send up a drone filming our work. Can you look at the drone afterwards. Report is always needed after each assignment. Time report, which resources

being used etc..”

Participant 6 5.1.3 Search

In the scenario, when it turns out that this is a drowning accident and that at least one person must have fallen off the boat, a search is initiated. JRCC has a system where you can enter information such as how much it blows, how current the water is, the size of the boat, etc. The system can then calculate where the person may be now and the search can begin at that area. Time is a vital factor in how large a search area can be. The longer time has passed, the greater the area where the person is likely to be.

One problem that was identified was that the search area is usually very large and it takes a long time to search the whole area. Helicopters can sometimes be used in searches, they usually contribute a lot because they can cover large areas very quickly, but the problem is that it takes them time to get to the site. The helicopters are equipped with thermal cameras which makes it easier to find people in large areas from afar. UxV could replace the helicopter here. The advantage with UxV is that the rescue personnel can bring it with them which lead to that it can be used in the search in the very beginning, which is rarely the case for the helicopter. It would also be possible to have several in the air at the same time. However, some participants pointed out that helicopter drivers do not want drones in the air while they are up, due to safety reasons.

“Then it is damn good to look for people on islands, so it is good with missing people in the archipelago environment so they take time even though we may have search dogs, it takes time to search every place. Should one be able to fly over and if someone does not hide with diligence, it is a

time gain in the archipelago environment to go ashore on each archipelago island and look and scream It takes time.”

Participant 8 5.1.4 Recipient of the information

Swedish maritime rescue mission can be seen as a hierarchical organization where the

information is distributed in a centralized coordinated way, where JRCC gives orders to OSC which in turn gives orders to the rescuer on site. Although it is JRCC that coordinates and controls the operation, it does not exclude that information between rescuer can be distributed in a decentralized way. If SSRS has any valuable information that they think the Swedish coast guard may need to know, they share it because they ultimately have a common goal with the mission.

Application of UxV during a rescue mission would mean an additional resource which can contribute information to the mission. This was a topic that was raised during the data collection and a question that was discussed a lot was, who should get the information from UxV It is clear from the analysis that JRCC should always be involved in information from the UxVs, they are the coordinator of the mission and can share the relevant information to

To drone, or not to drone : A qualitative study in how and when information from UxV should be distributed in rescue missions at sea