Visual Analysis of Military Diving Incident Reports

G. Walsh

, N. S. Andersen

, J. Kusnick

, E. B. Sørensen

and S. J

anicke

Centre for Visual Data Science, Department of Mathematics and Computer Science,

University of Southern Denmark, Denmark

Keywords:

Geospatial Temporal Visualization, Military Diving, Incident Reporting, Incident Analysis.

Abstract:

Military diving can be characterized by its specialized tasks, advanced equipment, and ever-changing opera-

tional environments, giving rise to remarkably unique operational challenges. Given this complexity, accidents

and incidents can occur. Incident reporting and analysis systems exist to collect data, perform trend analysis

on safety interventions, and modify unsafe behaviors. By examining existing Military Diving Incident Re-

porting (MDIR) systems and literature, we reveal that individual European countries have segregated systems,

but these lack standardization and interoperability. This paper introduces a novel visualization tool, focusing

primarily on rebreather incidents, a critical piece of equipment with a history of incidents in both military and

civilian contexts. We compare our proposed system with existing models, highlighting strengths and areas

for improvement. This tool aims to illustrate the potential of a broader, more comprehensive system, which

would cover not only rebreather incidents but all types of military diving incidents. The paper concludes with

insights into the potential of a comprehensive, standardized MDIR system, proposing future extensions and

research opportunities to enhance military diving safety and operational effectiveness.

1 INTRODUCTION

Despite being generally safe due to well-established

diving procedures and the expertise of trained per-

sonnel, military diving operates in a high-risk and

complex environment, making accidents a possibil-

ity (Breitstein and Nevo, 1999; O’Connor et al.,

2007). The diverse nature of military diving tasks, the

use of advanced equipment, and varying operational

contexts contribute to this risk. Unlike other high-

reliability military sectors like aviation, diving lacks

extensive safety research focused on human error.

The existing literature primarily focuses on equip-

ment design and impacts on cognitive and psychomo-

tor performance, primarily in the context of recre-

ational diving (Breitstein and Nevo, 1999; O’Connor

et al., 2007). To bridge this gap, this paper introduces

a prototype visualization tool for reporting military

diving incidents. This tool is intended to serve as

the basis for a more comprehensive system, enabling

divers to learn from past incidents through detailed

https://orcid.org/0000-0002-1095-5471

https://orcid.org/0000-0001-6926-1397

https://orcid.org/0000-0002-1653-6614

https://orcid.org/0009-0001-7915-957X

https://orcid.org/0000-0001-9353-5212

visual analysis of historical data.

This approach aligns with the broader goal of inte-

grating incident systems into military operational pro-

cedures, enhancing strategic planning and decision-

making. This is exempliﬁed by the ”As Low As Rea-

sonably Achievable” (ALARA) and ”Report, Ana-

lyze, and Get Better” (RAG) models typically cited

in military manuals (Department of the Navy, 2020;

U.S. Army, 1998). The ALARA model aims to re-

duce safety risks through controls designed to prevent

incidents and catch errors, ensuring no single control

becomes a point of failure. Similarly, the RAG cy-

cle emphasizes continuous collection and analysis of

reports to assess and improve risk control. This fa-

cilitates informed decisions based on observations of

safety incidents and hazards.

Regarding the development of our visualization

tool, Section 2 will delve into the complexities and

challenges of military diving. Understanding these

challenges is crucial for creating effective incident re-

porting systems in the sub-sea domain. Within this

section, we dissect four key aspects of military div-

ing, highlighting the critical information that should

be recorded at the data entry stage of a Military Div-

ing Incident Reporting (MDIR) system. Such infor-

mation is central for later analysis and visual repre-

sentation of insights to end-users. Additionally, we

Walsh, G., Andersen, N., Kusnick, J., Sørensen, E. and Jänicke, S.

Visual Analysis of Military Diving Incident Reports.

DOI: 10.5220/0012577700003660

Paper published under CC license (CC BY-NC-ND 4.0)

In Proceedings of the 19th International Joint Conference on Computer Vision, Imaging and Computer Graphics Theory and Applications (VISIGRAPP 2024) - Volume 1: GRAPP, HUCAPP

and IVAPP, pages 749-758

ISBN: 978-989-758-679-8; ISSN: 2184-4321

749

Figure 1: The sub-sea ecosystem of interconnected smart

diving devices and the potential challenges in achieving sit-

uational awareness during execution as well during the as-

sessment of such operations. Source: (CUIIS, 2023).

Figure 2: The usage of GeoTime as a geospatial analysis

software, allowing for the visual analysis of events over

time in 3D. Source: (Kapler and Wright, 2005).

discuss the technical requirements for an MDIR sys-

tem and its accompanying visualization tool, focusing

on aspects such as data integration, visualization ca-

pabilities, and the scalability of such systems.

In exploring the prerequisites for an all-

encompassing MDIR system, we review existing

systems, focusing on their strengths and weaknesses,

particularly those utilized within the EU and NATO.

Examining two state-of-the-art MDIR systems (Sec-

tion 3), we lay the groundwork for introducing our

prototype in Section 4, called the ’Rebreather Fatality

Monitor (RFM)’. While primarily educational, the

RFM represents a notable step towards a more

inclusive system that covers all types of military

diving incidents. Finally, Section 6 and Section 7

conclude on the prototype and present our vision

for a uniﬁed and comprehensive MDIR that aims to

surpass the capabilities of current systems.

2 DIVING DATA & MDIR

SYSTEM REQUIREMENTS

Military diving, encompassing a range of complex un-

derwater operations from defensive to offensive activ-

ities (Walsh et al., 2023b; J

zwiak et al., 2016), relies

on divers using breathing equipment independent of

any surface air supply. Given the intricacies of these

operations, MDIR systems must capture various data

types that reﬂect the unique challenges of military

diving. The following four key aspects highlight the

critical information necessary for input into an MDIR

system for subsequent visualization purposes.

1 Specialized Tasks. Military Diving Operations

(MDOs) involve diverse tasks like maintenance, con-

struction, search and recovery, ordnance disposal, and

tactical activities (US Navy, 2006b). In an MDIR sys-

tem, it is crucial to have detailed ”specialized tasks”

data entries. This granular data enables nuanced anal-

ysis by task, revealing correlations in safety, opera-

tional performance, and protocol adherence (Depart-

ment of the Navy, 2020). Advanced visualization

techniques are imperative in this aspect, aiding in pin-

pointing areas needing safety improvements and op-

timizing resource allocation. Task-based data cate-

gorization in MDIR systems simpliﬁes data ﬁltering

and enhances deviation detection, crucial for in-depth

analysis of each speciﬁc task in MDOs (Lock, 2023).

2 Advanced Equipment. In MDOs, increasingly

sophisticated equipment plays a crucial role in the

planning, executing, and assessing sub-sea opera-

tions. Equipment ranging from dive computers to

safety gear like smart hyperbaric chambers for health

monitoring and intervention (Kirwan et al., 1997; Lib-

eratore, 1998; CUIIS, 2023), are integral to oper-

ations. Therefore, a comprehensive MDIR system

must include standardized data entries for this equip-

ment. Such data is important for evaluating the equip-

ment’s impact on diver performance and for corre-

lating different data entries to identify patterns. Ad-

ditionally, during MDOs the sub-sea ecosystem can

consist of a network of interconnected equipment

and smart diving devices, as illustrated in Figure 1.

This illustration not only showcases the complexity of

the ecosystem but also emphasizes the challenges in

achieving interoperability among diverse equipment

types. It further highlights the need for a standardized

method to effectively fuse and present this data.

3 Environmental Conditions. Military divers

operate in challenging sub-sea conditions where en-

vironmental factors like depth, temperature, salinity,

and visibility play critical roles. These environmen-

tal factors can drastically inﬂuence diver performance

and equipment functionality (Walsh et al., 2023a).

Additionally, interactions with marine life and nav-

igation around underwater obstructions are signiﬁ-

IVAPP 2024 - 15th International Conference on Information Visualization Theory and Applications

750

Table 1: High-level MDIR system requirements, focusing on data collection, integration, and visualization components in

military diving, emphasizing scalability and continuous improvement for adaptability to evolving needs and practices.

Category Requirements Details

Data Collection Specialized Tasks Type of Task

Advanced Equipment Status and Metrics of Connected Items

Environmental Context Metrics and Notes

Data Integration Efﬁcient Integration Common Dataset

Data Validation Ensure Information Accuracy

Visualization Intuitive UI Geo-Spatial and Temporal Exploration

Visualization tool Features Support Individual Incident Summaries/Export

Training Design Suitable for Training Settings

Scalability Visualization tool Adaptability Accommodate varying volumes of MDIRs

Continuous Improvement Ongoing Evaluation Based on user feedback and evolving needs

cant hazards (Terrill, 2001). As such, potential users

of an MDIR system would be interested in captur-

ing such environmental factors and data for a more

accurate representation of events and risk mitigation

strategy. Such parameters could help support and aid

the reﬁnement of training programs, develop targeted

preventive measures, and design equipment that per-

forms robustly under various conditions.

4 Geospatial & Temporal Aspects. Incorporat-

ing geospatial and temporal data is crucial for so-

phisticated data visualization and analysis in MDIR

systems. The integration of geospatial data such as

latitude, longitude, and depth, along with temporal

data encompassing time and duration of operations,

enhances the understanding of where and when inci-

dents occur. This integration is important in identi-

fying trends and patterns over time, enabling future

operations to be strategically planned with an aware-

ness of historically problematic locations or times. By

providing additional context to incidents, geospatial

and temporal data can signiﬁcantly improve the un-

derstanding of incident dynamics, particularly in re-

lation to the aspects outlined in 1 – 3 . Further-

more, analyzing the locations and times of past inci-

dents allows for the tailoring of future military diving

operations, addressing speciﬁc challenges and scenar-

ios that divers are most likely to encounter. The use

of tools like GeoTime, seen in Figure 2, which offers

a combined temporal and geospatial display, further

demonstrates the potential of such integration in vi-

sualizing and tracking military operations effectively,

and similarly in the sub-sea operational environment

(Kapler and Wright, 2005).

MDIR System Requirements. The inherent chal-

lenges of MDOs and the criticality of capturing data

for visual analysis to reveal incident patterns were

described in 1 – 4 and are further detailed in

our high-level MDIR system requirements (Table 1).

These aspects primarily concern the effective collec-

tion and integration of data. A related requirement is

the seamless integration with standard datasets cou-

pled with robust validation methods to ensure data

accuracy. Also outlined in Table 1 is the need for

advanced visualization capabilities providing a mul-

tifaceted view of the data, enabling a thorough exam-

ination and insightful analysis, which in turn is in-

strumental in revealing incident patterns. In this re-

gard, a user-friendly interface that supports geospatial

and temporal exploration is also vital, offering an in-

tuitive and meaningful entry point for incident inves-

tigation. Furthermore, the MDIR visualization tool

should not only efﬁciently summarize and export in-

dividual incident reports for data sharing and interop-

erability purposes but also be adaptable for training

purposes. Scalability, too, is a critical factor, ensuring

the system can handle varying volumes of MDIRs ef-

fectively. The last, yet equally important aspect, is the

incorporation of a mechanism for ongoing evaluation

and improvement, anchored in user feedback and the

dynamic needs of military diving operations. Such a

mechanism is imperative to sustain the system’s ef-

fectiveness and relevance. This approach, steered by

requirements, aims to establish a solid foundation for

visual analysis, enhancing safety, operational perfor-

mance, and resource optimization in military diving.

3 EXISTING MDIR SYSTEMS

Due to the restricted nature of current MDIR sys-

tems and related systems, for this paper, we chose

to contextualize our proposed MDIR system against

two tools that set out to solve similar problems: (i)

The Next-Generation Incident System (NICS) and (ii)

Jump Dive Reporting System (DJRS). The study of

these precedents aids our effort to extrapolate relevant

Visual Analysis of Military Diving Incident Reports

751

Figure 3: Next-Generation Incident Command System

(NICS), a web-based geospatially referenced Command

and Control system developed to enhance information shar-

ing, situational awareness, and decision support for the

emergency response community. Source: (Hogan and Fos-

ter, 2022).

features, characteristics, and functionality to be in-

cluded in our MDIR prototype visualization tool and

highlight areas where our prototype can provide addi-

tional functionality.

An examination of military reporting tools reveals

that while such tools exist, they are predominantly

classiﬁed due to their sensitive nature. Furthermore,

in the case of the European market, it is evident that

such tools, as with other military products and sys-

tems, are country-speciﬁc, lacking interoperability,

and have suffered from years of under-investment and

under-development (Von der Leyen, 2023).

Despite the clear advantages of collaborative de-

velopment, European joint defense research and tech-

nology investment in 2022 constituted only 18 per-

cent of the total spending, a ﬁgure substantially lower

than the European Union’s target of 35 percent, as re-

ported by the European Defence Agency (European

Defence Agency, 2023). The increase in collabora-

tive defense research and technological investment is

crucial. It would not only enable member states to

contribute more effectively to the development of ad-

vanced military tools but also reap the beneﬁts from

the enhanced insights and capabilities these tools pro-

vide (Von der Leyen, 2023).

3.1 Next-Generation Incident

Command System

Developed by the MIT Lincoln Laboratory, the Next-

Generation Incident Command System (NICS) is a pi-

oneering web-based system designed to enhance in-

formation sharing and situational awareness among

emergency response teams. It can be seen in Figure 3

(Hogan and Foster, 2022). The advantages of this

system are the geospatially referencing, used for ac-

curate location tracking, and enabling robust person-

nel tracking to accurately record disaster responses

(Tarakanov, 2023; Hogan and Foster, 2022). Simi-

Figure 4: The Dive Jump Reporting System (DJRS), a tool

examining safety in parachute jumps and underwater dives.

Users can access malfunction reports, evaluate new equip-

ment, and explore data. Users can examine reports, log

numbers, dive dates, types, purposes, statuses, and respon-

sible personnel. Source: (Naval Safety Command, 2018).

larly, this tool also allows for near real-time sharing of

diverse data types. These characteristics share some

parallels with what could aid MDIR systems. How-

ever, the NICS has some disadvantages and lacks an

intuitive user interface, the ability to gain a compre-

hensive overview of multiple incidents, and advanced

ﬁltering of such events.

3.2 Dive Jump Reporting System

When looking at military reporting systems, we

brieﬂy looked at many examples, including US CALL

(Center for Army Lessons Learned) (US Army,

2023), and U.S. Air Force Safety Automated Sys-

tem (AFSAS) Preliminary Mishap/Incident Reporting

(US Air Force, 2023) among others. Ultimately, the

Dive Jump Reporting System (DJRS) is a military re-

porting tool used by the U.S. military, focusing on

safety in parachute jumps and underwater dives (US

Navy, 2006a). It resonated most clearly with the goals

of our prototype MDIR visualization tool. We thus

examined it in further detail as a precedent.

As shown in Figure 4, DJRS is similarly lack-

ing in the usage of advanced visualization techniques,

which distill and extrapolate the critical information

and history regarding such operations, which is easy

to consume by the end-user. The tool enables users

to review malfunctions and diving mishaps, exam-

ine new research and development equipment perfor-

mance, and explore incident reports. However, while

this data is available in the portal, it is difﬁcult to ﬁnd,

cumbersome to go through, and extremely text-heavy.

This is similarly true for the type and implementa-

tion of the DJRS system. Thus, a design considera-

tion and requirement of our MDIR prototype is that

it should utilize data visualization techniques that en-

sure ease of access to historical operational data in an

overview plus details in an on-demand format. Simi-

larly, the design should enable the fundamental anal-

IVAPP 2024 - 15th International Conference on Information Visualization Theory and Applications

752

ysis of incident reports to identify trends more eas-

ily, perform geospatial analysis, perform equipment

evaluation, and spot safety correlations. Furthermore,

the functionality to easily ﬁlter displayed results such

as dive types, dates, depth, and so forth would make

the tool much more valuable to the end-user. Another

weakness of this and similar systems is their lack of

ability to be utilized cross-organizationally, reducing

their effectiveness given the bespoke and limited fre-

quency of such operations.

4 PROTOTYPING THE MDIR

VISUALIZATION TOOL

Our proposed prototype focuses on addressing points

and requirements raised in Section 2 by creating an

interactive visualization tool. The prototype focuses

on rebreather data ( 2 ), due to data availability on

rebreather fatalities for both the military and civilian

contexts, using a dataset published and maintained

by Deep Life Group (Deep Life, 2012), listing re-

breather fatalities with corresponding details on the

incidents. This dataset helps demonstrate our proof

of concept but has shortcomings. Those include a rel-

atively unstructured data format, as individual users

originally input data without a strict template. Simi-

larly, geospatial information in the dataset varies sub-

stantially, so locations are approximated based on the

information available.

For the development of our prototype, we se-

lected Dash by Plotly (Plotly, 2023) as our frame-

work, complemented by Plotly for creating interac-

tive visualizations, as illustrated in Figure 5. Our so-

lution is grounded in the Information Seeking Mantra

”Overview ﬁrst, zoom and ﬁlter, then details on de-

mand” (Shneiderman, 1996). This approach gives

users an overview of all rebreather fatalities and al-

lows user interactions to drill down on the designated

patterns of interest. Within our visualization tool, var-

ious highlighting options are available, aligned with

the four key aspects identiﬁed in Section 2. This fea-

ture is designed to present incidents in a way that

not only meets the criteria set out in Table 1 but also

addresses the limitations of existing solutions, as de-

tailed in Section 3 and Table 2.

Our visualization tool is organized ﬁrstly with an

overview, aiming to provide the user with a good

grasp of where and when incidents occurred using the

Incident Map and the Incident Timeline, accompa-

nied by further Statistic Visualizations. These visu-

alizations are linked by mouse Interactions, allowing

the user to zoom, ﬁlter, and demand detailed informa-

tion. The constituent components of the visualization

tool are described in the following:

Incident Map ( 4 ). The ﬁrst and most signiﬁ-

cant part of the visualization tool (see Figure 5 (A))

is the interactive Incident Map using Mapbox (Map-

box, 2006). Each incident is represented by a geo-

located circular dot colored in a shade of blue based

on the depth of the incident (overview). The color

scale is shown to the left of the map. Whereby a

gray and smaller dot depicts dives of unknown depths.

Dots overlap in the same position, which introduces

a slightly disorganized view for users and could be

avoided in the next iteration by clustering the entries.

Adding additional map layer functionality and a de-

tailed maritime map could also be included.

Incident Timeline ( 4 ). The temporal representa-

tion of the incidents consists of two parts: A blue-

colored area chart pointing downwards to depict the

depth of the dives throughout the years (see Figure 5

(B)). This chart starts at the sea level, represented as

zero on the y-axis, and extends negatively towards

the deepest recorded dive, thus visually mirroring the

descent to the sea ﬂoor. A slider to the left of the

chart allows the user to adjust the highlighting selec-

tion based on the depth. Beneath it is the stacked area

chart (see Figure 5 (C)) colored according to expe-

rience levels of divers in different shades of green.

We decided against a combined, single timeline to

highlight the importance of the actual depth, which

is negative, compared to the positive values of inci-

dent numbers. A combined timeline would further-

more be a complex visualization for users to decode

and interpret. Instead, the two area charts are aligned

and linked by interactions to show the correlations be-

tween the different data dimensions. Similarly, the

period in which the dive took place is selectable us-

ing the Incident Timeline, differentiated by experi-

ence level (see Figure 5 (C), highlighting the number

of incidents that occur within a particular time range.

Further Statistics ( 1 , 3 ). To allow for a concise

overview of basic descriptive statistics, two stacked

bar charts are introduced to get an overview of the dis-

tribution of age and dive type (Figure 5 (D)). The col-

ors are kept in line with the Incident Timeline and re-

ﬂect the divers’ experience level. For that purpose, the

ﬁlled area plot combined with the depth range slider

(Figure 5 (B)) helps to highlight rebreather fatalities

based on depth.

Interactions. The visualization tool consisting of var-

ious juxtaposed but interlinked visualizations is based

on the idea of multiple-linked views (Wang Baldon-

ado et al., 2000), especially the combination of tem-

poral and geospatial visualizations (Incident Map and

Timeline) (Andrienko and Andrienko, 2006). All vi-

sualizations display the same (sub)set of data points,

Visual Analysis of Military Diving Incident Reports

753

Table 2: Comparative Evaluation of MDIR System versus NICS and JDRS based MDIR system requirements, as well as

additional criteria. The evaluation demonstrates how the proposed MDIR performs concerning the stated requirements. The

evaluation reveals opportunities for further improvement in terms of data collection and integration, data accuracy, scalability,

and real-time data incorporation.

Criteria Proposed MDIR System NICS JDRS

Data Collection and Integration Moderate Moderate Moderate

Data Visualization Excellent Moderate Moderate

User Interface and Experience User-Friendly Non-Intuitive Moderate

Scalability and Adaptability Moderate Moderate Limited

Data Accuracy and Validation Moderate Moderate Moderate

Real-Time Data N/A N/A N/A

Task-Speciﬁc Analysis Moderate Basic Basic

Environmental Context Integration Moderate Basic Basic

Training and Improvement Focus Strong Weak Weak

Figure 5: MDIR visualization tool screenshot illustrating the interactive incident map (A), which displays the distribution

of diving incidents by depth geospatially, incident depth ﬁlter (B), incident timeline (C), and statistical visualizations (D).

(C) uses a ﬁlterable stacked area chart to communicate to the user the distribution of incidents over time, broken down by

experience level. Accompanying stacked bar charts (D) breaks down incidents by diver expertise, age group, and the type and

timing of dives, providing a multifaceted view of diving trends and safety metrics at a macro level.

IVAPP 2024 - 15th International Conference on Information Visualization Theory and Applications

754

and the highlighting interactions are applied to all

of them. So, when initializing the tool, the whole

dataset is visualized, and the various data dimensions

offer opportunities to ﬁlter down the selection of high-

lighted data entries. We implement a Focus+context

method (Card et al., 1999) to highlight speciﬁc inci-

dents the user might be interested in while keeping the

valuable information of the context around by high-

lighting the selected ones and graying out the others.

The map allows narrowing down the dataset through

geospatial selection of interesting areas by zooming

into the map. In parallel, the timelines highlight the

inspected period in time, the depth, or the level of ex-

pertise of divers with accidents.

For example, suppose an age bracket or dive group

type is selected in the stacked bar charts. In that case,

the prototype highlights the corresponding data points

by displaying them in yellow on the incident time-

line and statistical visualizations and displays the un-

selected incidents on the map as gray. By doing this

and displaying all the incidents on the incident map,

the geospatial context is maintained while highlight-

ing the important features.

In addition to the Incident Map, the Incident Time-

line, and Statistic Visualizations, our MDIR prototype

enables the user to view detailed and comprehensive

information regarding a speciﬁc incident.

When an incident is selected, a pop-up window

displays the complete information, including name,

age, rebreather type, country, coordinates, depth,

summary, training implications, design implications,

and sources. E.g., an incident that occurred in Por-

tugal as part of a training dive with an expert-level

scuba diver appears to be a Belgium Military Diver

taking part in a NATO exercise. Furthermore, a user

can understand where the incident occurred (geo-

located), at what depth, and the speciﬁc rebreather in-

volved in the fatality (in this case, a Semi-Closed Re-

breather). Additionally, the complete summary infor-

mation for each incident is available. In the example,

Belgian divers tended to remove the SCR PPO

mon-

itor from their equipment because they caused many

false alarms, resulting in a fatality in this incident.

5 DISCUSSION

Prototype Limitations. As this paper primarily fo-

cuses on the rebreather aspect of military diving, we

acknowledge that this represents only a fraction of

the diverse scenarios and equipment utilized in mil-

itary diving operations. The complexity of military

diving extends far beyond rebreathers, encompassing

a wide range of equipment, environments, and oper-

ational contexts as described in Section 2. Each of

these elements brings its own set of challenges and

risks. Recognizing this limitation, our prototype does

not support a wide range of equipment, environments,

and operational contexts in this iteration.

Another limitation of this prototype is the dataset

used. The dataset from (Deep Life Group, 2020), used

as a proof of concept, is lacking in accuracy in terms

of data formatting and precise geospatial and some-

times temporal data entries. This has resulted in an

approximation of some location markers in the proto-

type. Similarly, the dataset contains a lot of unknown

or blank entries, particularly relating to training and

design implications. This, in turn, has a negative ef-

fect on the use of this tool for training and design pur-

poses. However, this point highlights a gap and need

for a more robust method of reporting and collect-

ing information about rebreather fatalities and mili-

tary diving incidents more broadly.

Prototype Insights. The intuitive highlighting in-

teractions included in the prototype, as discussed in

Section 4, make it possible to perform an exploratory

search of incidents geospatially according to several

criteria mentioned in Section 4. This aspect is novel to

this visualization tool, as seen in Table 2 where other

tools do not implement such features.

6 CHALLENGES &

OPPORTUNITIES

Challenges. The development of the proposed

MDIR system has presented different challenges.

Speciﬁcally in the context of military diving, the secu-

rity of divers and mission-related data are under strict

scrutiny regarding who receives access to operational

and dive data and for what purposes. Development

of the proposed MDIR system would require support

from interested parties and their buy-in regarding the

sharing of operational information with other defense

forces. This would be a signiﬁcant challenge to over-

come. However, it is not without its many semiotic

beneﬁts. For instance, without mutual information

sharing across different forces, the volume of data

collected from national diving operations might be

too limited to discern patterns or analyze trends effec-

tively. This data is vital for enhancing diver training

and preparation. Overcoming these challenges is fea-

sible through responsible and balanced data-sharing

practices. A transparent and secure approach to data

exchange in any future MDIR system is essential, not

only for the system’s functionality but also for main-

taining the trust and safety of all involved parties.

Visual Analysis of Military Diving Incident Reports

755

Opportunities. The prototype revealed signiﬁcant

improvements that can be made to improve the status

quo of European and NATO MDIR systems. A stan-

dard method of recording data in such MDOs would

need to be implemented to establish a common MDIR

system. The ultimate aim of a common MDIR system

is to overcome national boundaries and operational si-

los and instead offer increased sharing of information

and insights among military entities, particularly re-

lating to historical data of incidents. This could be

seen in a similar light to what currently exists in the

aviation sector with (European Union Aviation Safety

Agency, 2024). As mentioned in Section 5, there are

criteria as seen in Table 2 still to be fully realized in a

comprehensive MDIR system. As such, there is am-

ple opportunity to further increase the knowledge and

insight generation of the tool according to the sys-

tem requirements. This could cover a more exten-

sive evaluation of equipment performance, a deeper

understanding of environmental risks and conditions,

and how this correlates to the type of dive operation

through the MDIR tool.

6.1 Extending the Prototype

A future extension or iteration of a European MDIR

system could incorporate Navigation Systems, Com-

munication Systems, Thermal Protection, Thermal

Imaging, Sonar Systems, Heads-Up Displays, Under-

water Drones and ROVs, Diver Propulsion Vehicles,

Hyperbaric Chambers, and Biometric Systems to ad-

dress unrealized opportunities. For example, if real-

time data were available, an enhanced MDIR system

could build on the prototype by including advanced

analytics to predict potential mishaps based on histor-

ical data and current conditions. A further possibil-

ity of extension could lie in collaborating with equip-

ment manufacturers to integrate equipment-speciﬁc

data streams to identify equipment malfunctions or

potential design ﬂaws better. A thoroughly devel-

oped extension could cover the range of diving equip-

ment and scenarios we could not address in our pro-

totype, providing a more representative view of and

insights into military diving. However, such an exten-

sion would still require a comprehensive and accurate

data collection similar to that mentioned in Table 2,

such that advanced data analysis techniques can be

implemented to increase analysis capabilities.

For example, in an extended MDIR system, a mili-

tary diver may want to deduce whether their dive pro-

ﬁle for their ascent or descent was appropriate. The

concept with this broader MDIR proposal is that they

could examine this using these data points in conjunc-

tion with safety system bio-metrics such as O

, CO

and heart rate to establish if it was safe and optimal.

6.2 Further Research Opportunities

The establishment of a common European MDIR sys-

tem opens the door to numerous research areas and

opportunities, including two novel opportunities:

Collaborative Research. The system can serve as

a platform for collaborative research projects

among NATO and European nations, pooling re-

sources and expertise to encourage and promote

enhanced innovation.

Predictive Modeling. Utilizing advanced analytics,

a future iteration of the MDIR system could

be used to develop predictive models for diving

mishaps, offering proactive solutions (with real-

time information integration).

The future potential of such a comprehensive and

common MDIR system is vast. It not only promises

enhanced safety and understanding of diving inci-

dents but also fosters a culture of collaboration, inno-

vation, and continuous improvement to better support

the challenges faced by divers in action.

7 CONCLUSION

This paper highlights the urgent need for a uniﬁed

MDIR system in Europe. Limited by national bound-

aries, current systems restrict collective learning from

incidents and near misses, impacting training, safety,

and risk assessment.

The development of a comprehensive European

MDIR system, as conceptualized through the proto-

type in Section 4, holds signiﬁcant promise for the

future of military diving safety and analysis. Beyond

data visualization, this paper lays the groundwork for

a platform that integrates data collection, analysis,

and visualization, identifying key areas for improve-

ment such as scalability and data accuracy (Table 2).

The developed visualization tool allows for the

global, geospatial, and temporal exploration of re-

breather fatalities in both civilian and military con-

texts. It facilitates incident analysis by highlighting

key factors like type, location, depth, and year and

uses graphs to easily identify trends in fatalities.

A comprehensive European MDIR system, as en-

visioned in this paper, aspires to be more than just

a tool for visualization and analysis; it aims to rede-

ﬁne and elevate the standards for a uniﬁed European

approach to military diving safety. While currently

focused on the rebreather aspect, the tool in this pa-

per lays the groundwork for a more expansive sys-

tem that can integrate various facets of military div-

ing, ranging from Navigation Systems to Biometric

IVAPP 2024 - 15th International Conference on Information Visualization Theory and Applications

756

Systems. Such a system hopes to promote and foster

joint research endeavors among NATO and European

nations, leveraging collective expertise to drive inno-

vation in various relevant research domains.

ACKNOWLEDGEMENTS

The survey paper includes some images under the fair

use copyright principle for commentary and discus-

sion purposes (image courtesy of (Hogan and Fos-

ter, 2022; Naval Safety Command, 2018). The Eu-

ropean Defence Industrial Development Programme

has supported this research under the project CUIIS

(Grant Agreement EDIDP-UCCRS-EDD-2020-059

— CUIIS). The authors independently produced this

work, which does not reﬂect the European Commis-

sion’s views, nor is the Commission responsible for

any use of the paper’s information.

REFERENCES

Andrienko, N. and Andrienko, G. (2006). A Framework for

Using Coordinated Displays for the Analysis of Mul-

tidimensional, Spatial, and Temporal Data. In Geo-

graphic Hypermedia, pages 293–308. Springer.

Breitstein, S. and Nevo, B. (1999). Psychological and be-

havioral aspects of diving. Best Publishing Co.

Card, S. K., Mackinlay, J., and Shneiderman, B. (1999).

Readings in information visualization: using vision to

think. Morgan Kaufmann.

CUIIS (2023). Comprehensive underwater intervention in-

formation system. URL: https://cuiis.eu/scope. Vis-

ited: 2023-09-09.

Deep Life (2012). Example iec 61508 safety case: Learning

from accident data. URL: https://deeplife.co.uk/or a

ccident.php. Visited: 2023-09-09.

Deep Life Group (2020). Iec 61508 safety case: Learning

from accident data. URL: https://www.opensafetygl

obal.com/or accident.php. Visited: 2023-09-09.

Department of the Navy (2020). Navy Safety and Occupa-

tional Health Manual. URL: https://secnav.navy.mil/

doni/SECNAV%20Manuals1/5100.23%20CH-2.pdf.

Visited: 2024-01-15.

European Defence Agency (2023). Defence data 2022: Key

ﬁndings and analysis. URL: https://eda.europa.eu/do

cs/default-source/brochures/2022-eda defencedata w

eb.pdf. Visited: 2023-09-09.

European Union Aviation Safety Agency (2024). Easa —

european aviation safety agency. URL: https://www.

easa.europa.eu/en. Visited: 2024-01-15.

Hogan, G. and Foster, S. (2022). The next-generation inci-

dent command system (nics). In Enhancing Capabil-

Dissemination website available at: https://cuiis.eu/

ities for Crisis Management and Disaster Response,

pages 23–32. Springer.

zwiak, D., Siermontowski, P., Da¸browieckiI, Z., and Ol-

sza

nski, R. (2016). Analysis of the risk of diving ac-

cidents in military and recreational diving. Polish Hy-

perbaric Research, 53(4):–.

Kapler, T. and Wright, W. (2005). Geotime information vi-

sualization. Information visualization, 4(2):136–146.

Kirwan, B., Basra, G., and Taylor-Adams, S. (1997). Core-

data: a computerised human error database for hu-

man reliability support. In Proceedings of the 1997

IEEE Sixth Conference on Human Factors and Power

Plants, 1997. ’Global Perspectives of Human Factors

in Power Generation’, pages 9/7–912.

Liberatore, T. C. (1998). Risk analysis and management of

diving operations: Assessing human factors. PhD the-

sis, Monterey California. Naval Postgraduate School.

Lock, G. (2023). Learning review for diving.

Mapbox (2006). Maps, navigation, search, and data. URL:

https://www.mapbox.com/.

Naval Safety Command (2018). Risk management infor-

mation - dive and jump reporting system (rmi-djrs).

URL: https://navalsafetycommand.navy.mil/Portals/

29/Documents/RMI-DJRS%20Smart%20Sheet.pdf.

Visited: 2023-08-23.

O’Connor, P., O’Dea, A., and Melton, J. (2007). A method-

ology for identifying human error in u.s. navy diving

accidents. Human Factors, 49(2):214–226.

Plotly (2023). Plotly: The front end for ml and data science

models. URL: https://plotly.com/dash/. Visited:

2023-09-09.

Shneiderman, B. (1996). The eyes have it: a task by data

type taxonomy for information visualizations. In Pro-

ceedings 1996 IEEE Symposium on Visual Languages,

pages 336–343.

Tarakanov, A. (2023). Ofﬁcial maintainer of next-

generation incident command system (nics). URL:

https://www.kontur.io/portfolio/nics/. Visited: 2024-

01-15.

Terrill, C. (2001). Romancing the bomb: Marine animals in

naval strategic defense. Organization & Environment,

14(1):105–113.

US Air Force (2023). Air force safety automated system

(afsas) preliminary mishap/incident reporting. URL:

https://afsas.safety.af.mil/. Visited: 2023-09-09.

U.S. Army (1998). Risk management [ﬁeld manual fm 100-

14]. URL: https://apps.dtic.mil/sti/pdfs/ADA402771

.pdf. Visited: 2023-08-23.

US Army (2023). Center for army lessens learned - driving

army change. URL: https://www.army.mil/CALL.

Visited: 2023-09-09.

US Navy (2006a). Joint deﬁciency reporting system. URL:

https://www.jdrs.mil/. Visited: 2024-01-15.

US Navy (2006b). Us navy diving manual, 6th revision.

URL: www.navsea.navy.mil/Portals/103/Document

s/SUPSALV/Diving/Dive%20Manual%20Rev%206

%20with%20Chg%20A.pdf. Visited: 2023-08-23.

Von der Leyen, U. (2023). President von der leyen makes

call for powering up european defence. Visited: 2024-

01-15.

Visual Analysis of Military Diving Incident Reports

757

Walsh, G., Andersen, N. S., J

anicke, S., and Stoianov, N.

(2023a). The lack of specialized symbology and vi-

sual interaction design guidance for sub-sea military

operations. In VisGap - The Gap between Visualiza-

tion Research and Visualization Software, pages 19–

26. The Eurographics Association.

Walsh, G., Andersen, N. S., Stoianov, N., and J

anicke,

S. (2023b). A survey of geospatial-temporal visual-

izations for military operations. In Hurter, C., Pur-

chase, H., and Bouatouch, K., editors, Proceedings

of the 18th International Joint Conference on Com-

puter Vision, Imaging and Computer Graphics Theory

and Applications - IVAPP, volume 3, pages 115–129.

SCITEPRESS Digital Library.

Wang Baldonado, M. Q., Woodruff, A., and Kuchinsky, A.

(2000). Guidelines for Using Multiple Views in Infor-

mation Visualization. In Proceedings of the Working

Conference on Advanced Visual Interfaces, AVI ’00,

page 110–119, New York, NY, USA. Association for

Computing Machinery.

IVAPP 2024 - 15th International Conference on Information Visualization Theory and Applications

758