Review of U.S. Coast Guard Vessel Stability Regulations (2018)

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

55 were made, as identified by the operator or as found during the survey, one of the proposed methods of lightweight measurement can be performed. If no changes or only minor changes are confirmed, then there should be no need to carry out a measurement at that time. It is important that the operator and the OCMI are requested to consider the issue of passenger ship lightship weight periodically. 5. OTHER STABILITY STANDARDS, CASUALTY DATA, AND STABILITY GUIDANCE Other Stability Standards Applicable to Coastal and Inland Vessels The regulations in the CFR are the main stability regulations applied to approximately 70% of the total USCG-inspected passenger and cargo vessel fleet, which are primarily non-ocean- going, working in inland service or in protected waters (see Appendix B). Comparing the CFR regulations with other international, class, and industry stability standards for coastal and inland vessels is more relevant than ocean-going vessel standards that include IMO and SOLAS stability standards. The stability regulations for coastal and inland vessels are generally structured to avoid detailed calculations. As inland vessel operators do not have the same licensing requirements as blue water vessel operators, they need stability guidance that is appropriate for the level of required knowledge. Often, adequate stability is determined by simple operational boundaries (protected waters) or by design, such as draft marks, loading limits, stack height limits, or maximum number of passengers on the upper deck. As inland and coastal shipping are often limited to voyages in protected waters, the stability rules are set primarily by state or regional entities throughout the world. There is no overriding United Nations (UN) body, such as IMO, regulating this type of vessel trade. Features, such as ship type, operational zones, size, or open or closed deck, often structure the

56 rules. Since there are many national maritime regulations that exist throughout the world to describe and assess in detail, this section will focus on the continental European comprehensive regulations. Classification societies also set stability rules for inland vessels. Yet many inland vessels in Europe are not classed and are not required to abide by the rules of class and hold a classification certificate. Local authorities certify the vessels, either to their own standards or to regional standards. Industry standards also exist, mainly for small boats and pleasure craft. Some equipment such as cranes may be designed to meet industry standards such as those of the International Organization for Standardization (ISO), German Institute for Standardization (DIN), or European Conformity (CE). A discussion of these specific requirements for flags, states, classes, industry standards, and related research into physics-based stability standards are contained in Appendixes I and J. Dynamic-Motions-Based Stability Standards As modern computing capability has increased, the ability to calculate complex stability-related motions of ships in wind and waves is becoming more of a reality. This ability will allow a better understanding of ship dynamics in extreme conditions and provide a way to assess and optimize the effects of design changes on vessel survivability better in these conditions. It is currently the subject of a major initiative at the IMO to enhance the traditional and prescriptive 2008 Intact Stability Code with the development of the so-called second-generation ship stability criteria. This effort is in response to the fact that certain vessels that satisfy the existing prescriptive criteria may still be vulnerable to various capsizing mechanisms. Although the IMO has not formally adopted this methodology, a tremendous international effort has gone into proposing a rational and physics-based multitiered analysis methodology, which starts with the prescriptive

57 criteria (Level 1) and ends with a direct stability assessment using numerical simulation or physical model testing (Level 3). The most interesting and elusive aspect of this effort is the so- called Level 2 analysis procedure, which tries to bridge the gap between greatly simplified prescriptive methodologies and time-consuming and expensive numerical simulation and physical model testing. Much effort has gone into validating and tuning these approaches and it is continuing. These Level 2 procedures have the potential to incorporate much of the research that has gone on in the field over the last half century. It is essential that these efforts include accurate physical modeling, nonlinear dynamics, and stochastic dynamics. The Subcommittee on Ship Design and Construction’s (SDC’s) final report on the finalization of the second generation ship stability criteria indicated that direct dynamic stability assessment requires considerable time to perform complicated numerical simulation codes with expensive experimental work needed to validate the simulation codes (see IMO SDC, 2018). It is, therefore, questionable whether direct stability assessment can be considered practical alternative criteria or not. Notwithstanding these difficulties, the efforts to develop physics-based dynamic motion stability criteria that are more realistic than the current static criteria should continue. Such research may be particularly relevant as new types of ships and floating platforms are proposed and built. Appendix J provides an overview of the development of dynamic- motion-based stability analyses. Casualty Statistics and Data Casualty and near miss data and statistics are a key source of information on where problems are occurring with vessel stability and the seriousness of any problems. The committee was tasked with identifying candidate sources of data that USCG could explore to inform its assessments of the advantages and disadvantages of requiring all inspected passenger vessels to undergo

58 periodic stability inspections or lightweight surveys. Such data for stability purposes are not readily available. This lack of data is part of a larger issue about stability data availability and collection in general, resulting in incomplete and inconsistent data, including near miss data. This topic is considered important enough to include in this report. Current State of Casualty Statistics Related to Stability In the United States, accident and incident data for marine transportation analysis are stored in various systems and managed by different agencies and entities, including USCG, the National Oceanic and Atmospheric Administration (NOAA), various Marine Exchanges and port authorities, and state and local oversight and regulatory agencies, as well as environmental groups, citizen action groups, and government oversight groups. The data are not stored in a common repository that is accessible to those with requirements for data access, storage, or analysis, nor are the data stored in a common format, which would facilitate data validation, reconciliation, analysis, and reporting. Worldwide, vessel accident and incident data are stored and managed by government agencies or organizations under the supervision of national ministries or governmental departments. These data are reported, collected, and compiled over time in databases in accordance with national regulations (Mullai and Paulsson, 2011; Kalyvas, Kokkos, and Tzouraminis, 2017). In several countries, national organizations such as Germany’s Federal Bureau of Maritime Casualty Investigation (2018) and the Hellenic Marine Accident and Events Investigative Service (2018) provide information and statistics about marine accidents within their national jurisdiction. Open-source accident records are available from the U.S. National Transportation Safety Board (2018); UK Marine Accident Investigation Board (Marine Accident

59 Investigation Board, 2018); and the European Maritime Safety Agency (2018), as well as from IMO’s Global Integrated Shipping Information System (2018),18 among others. In the United States, marine casualty statistics related to stability are captured in the USCG MISLE database.19 Data fields for stability statistics generally fit into three categories— sinking, loss of stability, and capsizing. USCG issues annual reports to the PVA that provide an overview of safety-related statistics for passenger vessels in the U.S. fleet. In the 2016 report, 5,765 marine vessel events occurred to inspected passenger vessels between 2013 and 2017, yet stability-related casualties were a small percentage of overall accidents (see Table 4; emphasis added). Table 4 Vessel Events for Marine Casualties Involving Inspected Passenger Vessels (2013 thru September 2016) Event Type 2013 2014 2015 2016* Total Material Failure/Malfunction 817 861 576 212 2,466 Loss/Reduction of Vessel Propulsion/Steering 537 581 384 174 1,676 Grounding 145 115 96 31 387 Allision 90 75 60 43 268 Loss of Electrical Power 61 70 63 27 221 Discharge/Release—Pollution 45 68 48 16 177 Flooding—Initial 34 32 41 15 122 Vessel Maneuver 10 8 30 41 89 Collision 26 22 10 21 79 Fire—Initial 24 18 23 11 76 Fouling 19 19 16 5 59 Flooding—Progressive 12 8 11 4 35 Set Adrift 9 12 8 5 34 Sinking 6 7 4 5 22 Fire—Reflash 3 3 5 2 13 Abandonment 1 1 4 2 8 Wave Strikes/Impacts 1 2 4 7 Loss of Stability 2 2 1 1 6 18 For example, the IMO reporting formats for stability-related casualties, which feed data into IMO’s Global Integrated Shipping Information System, involve completion of a Damage Card and Intact Stability Casualty Records (see MSC-MEPC.3/Circ.3, ANNEX 5 form at http://www.imo.org/en/OurWork/MSAS/Casualties/Documents/MSC-MEPC.3-Circ.3.pdf). 19 The USCG Marine Accident report form CG-2692 is one of the data forms used to feed information into MISLE. It can be found at the following site: https://www.dco.uscg.mil/OurOrganization/AssistantCommandantforPreventionPolicy(CG- 5P)/InspectionsCompliance(CG-5PC)/OfficeofInvestigationsCasualtyAnalysis/2692ReportingFormsNVIC01- 15.aspx.

60 Vessel Yawl/Pitch/Roll/Heel 5 5 Explosion 1 3 1 5 Cargo/Fuel Transfer/Shift 4 4 Damage to Cargo 2 2 Capsize 1 1 2 Well Blowout 1 1 Personnel Casualty—Injury 1 1 Total 1,843 1,907 1,386 629 5,765 SOURCE: Table 7 from USCG-PVA Quality Partnership Draft Annual Report, October 2016. NOTE: * = Data for 2016 is through September. Gaps in Available Casualty Statistics Problems with data to support modeling and analysis in marine transportation are well- documented (see National Research Council, 1983, 1990, 1994, 1999, 2003; National Transportation Safety Board, 2002; U.S. Committee on the Marine Transportation System, 2013; Mazaheri, Montewka, Nisula, and Kujala, 2015), having grown with the proliferation of electronic data. Data for maritime accident and incident analysis have varying storage requirements, exist in various formats, and are gathered and collected from various agencies and individuals around the world, with varying degrees of compatibility (National Research Council, 2003; Ellis, 2011). Missing, inaccurate, and underreported maritime accident and incident data challenges have been reported for many years (Hassel, Asbjørnslett, and Hole, 2011). As a result, data validation, compatibility, integration, and harmonization are increasingly significant challenges in maritime data analysis and risk assessments. Data retrieval from databases and data warehouses can be cumbersome, and is often more difficult, since no standard reliable database for near miss reporting or exposure data has been developed in marine transportation. The U.S. Government Accountability Office, Congress, and the National Academies of Sciences, Engineering, and Medicine (the National Academies) have explored methods to improve the collection, representation, integration, and sharing of accident and incident data (National Research Council, 1994; U.S. Department of Homeland Security, 2008),

61 as is available in other countries and venues (Storgård, Erdogan, Lappalainen, and Tapaninen, 2012). Impacts of Data Challenges on Casualty Data Analysis In marine transportation, as in other domains, event analyses are constrained by the quality of the data gathered, the maturity of the associated reporting system, and the training and background of the investigator and reporter (who may not be the same person). Such constraints place limits on the adequacy and strength of analyses conducted with maritime safety data, limitations that have been characterized and analyzed extensively in reports prepared by the National Research Council, the National Transportation Safety Board, and the U.S. Government Accountability Office (National Research Council, 1983, 1990, 1994, 1999, 2003; National Transportation Safety Board, 1994). Analysis of marine casualty and stability-related accident and incident data requires significant time for data reconciliation and cross-validation across data sources to ensure that data records are accurate, the data capture the entire event of record, and data redundancy is minimized. Data reconciliation and cross-validation are particularly challenging when data records from one data source capture the initial part of an event of record (e.g., an initiating mechanical failure), while data records from another reporting agency, describing the same event, capture the initiating event as well as the series of cascading and related events (e.g., a vessel sinking, an eventual accident). Absent a standard incident and accident coding scheme, common data storage, common transmission formats, and a common data dictionary, which defines accidents, incidents, unusual events, and contributory situations, marine casualty and stability-related data analysis and data record reconciliation can require extensive, time-consuming steps. These steps can include the

62 review of available paper and electronic sources, additional searches to confirm the events, and requests for additional information, to ensure that the entire event chain was captured. Because marine casualty and stability datasets can be incomplete, conflicting, missing, or inaccurate, data validation and resolution may require search and compilation of data sources not only from maritime safety sources, but also from vessel, traffic, transit, meteorological, charting, and geographic sources, as well as from other national, regional or provincial, state, local, public, or private sources. Stability-related casualty data could benefit from a standard descriptive data dictionary used by all data-gathering organizations to codify events, perhaps aligned with international, state, local, and private data storage and organization standards for capturing maritime safety data. Automated data analytic methods can be used to identify discrepancies and missing data in USCG MISLE records through text and data mining, leveraging machine learning and classification techniques, and proposing automated updating and reconciliation of those records, with human supervision (Grabowski, Dorsey, and Wang, 2018). Needs for Additional Data Missing, unavailable, and inaccessible maritime accident and incident data have been identified as a persistent problem in maritime risk modeling and data analysis studies for many years (Grabowski, Merrick, Harrald, Mazzuchi, and van Dorp, 2000; Hassel, Asbjørnslett, and Hole, 2011). To address these needs, marine casualty and stability-related event analysis could be enhanced with the availability of environmental data, such as wind, ambient air temperature and air pressure, dew point, and relative humidity, as well as information about visibility, extreme weather, and impending weather. In ports and regions that have NOAA’s Physical

63 Oceanographic Real Time System (PORTS®)20 installed, PORTS® data could be a primary source for historical waterway environmental data, such as water level, current, relative humidity, dew point, and visibility. However, PORTS® data are not available in all U.S. ports and waterways. Vessel transit, water level, and environmental data from the U.S. Army Corps of Engineers (USACE) could be integrated into existing maritime casualty and stability databases, along with environmental data available from the National Weather Service. Automated Identification System (AIS) data have been utilized as an additional secondary data source to supplement missing environmental data in marine casualty and stability databases; however, the use of this additional data source, which stores data in nonstandard formats incompatible with MISLE, requires additional data validation, reconciliation, and merger activities. The lack of quality, localized environmental data in formats compatible with accident and incident data limits the quality and granularity of the accident analyses that can be undertaken and the strength of the resulting analytics. Future Opportunities Opportunities to improve accident and incident data have been identified, many of which could improve the accuracy, clarity, and harmonization of data and processes for stability calculations. Future data needs and analyses will require solutions to the data issues identified. Increasing sensorization of industrial equipment, systems, fluids, components, bulkheads, and structures will result in ever larger needs for Big Data storage, retrieval, and integration (Diaz-Casas et al., 2018). Big Data cleansing, validation, fusion, and analytics are increasingly important, and key performance indicator (KPI) monitoring and predictive analytics are already central to tracking 20 For additional information, see https://tidesandcurrents.noaa.gov/ports.html.

64 and anticipating system and vessel performance through the Industrial Internet of Things (IIoT) and Shipyard 4.0 activities (Blanco-Novoa et al., 2018; Diaz-Casas et al., 2018). Repairing algorithms (Chu et al., 2013) possibly may be used to automatically repair inconsistent or “dirty” data by identifying conflicting data; the conflicts could be used to clean the data by removing outliers and abnormal data. Data fusion for multisensor data (Khaleghi et al., 2013) may be applied, in addition to unsupervised learning methods such as data clustering, which can automatically form groups of similar data, to enable combining information from several sources to form a unified picture. Recent work has focused on automating identification, flagging, and repair of missing, incomplete, and conflicting accident and incident data records in the USCG MISLE database (Grabowski, Dorsey, and Wang, 2018). Accident and incident data will increasingly include unstructured data (pictures, sound, animations, videos, PDFs, “likes,” tweets, and social media postings), as well as traditional structured data (text, numbers, and algorithmic). Unstructured data are already found in accident and incident reports, case files, descriptions, and social media comments. Current analytical tools such as text mining and natural language processing (NLP) can assist in (1) extracting useful, structured information from reports, images, videos, and descriptions to complement the existing structured data; (2) detecting potential risk indicators from daily reports; and (3) performing sentiment analysis from social media comments. A further challenge in future accident analyses lies in the mix of data volumes from multiple data sources, such as accident data compared to incident data, which could mean performing analysis on an unbalanced data set. Several data mining processes (Anand et al., 2011), including up- and down-sampling with clustering, may prove helpful in the context of unbalanced data analysis. Another approach may include applying a cascade framework in the accident data analysis, which performs a series of data chain

65 classifications, with each step in the chain analysis making the remaining data more balanced. The move to open-source data sets, and the integration of open-source, public, and private data sets, will result in additional accident and incident data analysis challenges (Grabowski et al., 2000; Kalyvas, Kokkos, and Tzouraminis, 2017). Stability Guidance to Vessels Maintaining the intended vessel stability under varying weather conditions encountered on a voyage requires proper guidance for the operating crew. The traditional static stability requirements do not necessarily reflect the environment of the vessel under extreme conditions. Proper guidance to the crew can assist them in understanding the risks to stability and improve awareness of when the vessel stability could be at risk. Examples of potential stability risks that vessels face include: 1. Damage to the hull, 2. Flooding caused by loss of watertight or weathertight integrity, 3. Extreme wind and sea conditions, 4. Shifting of cargo, 5. Improperly stowed cargo by untrained shoreside staff, and 6. Inadequate knowledge of propulsion system limitations that could lead to loss of propulsion in extreme conditions. Smaller Vessels The information provided for smaller vessels, those without SOLAS certificates or stability booklets, may not be sufficient for extreme conditions. Additional information on extreme conditions, when stability may be at risk, would be useful to these vessels, and the committee

66 encourages USCG to work with industry advisory groups to see what guidance can be provided on board for vessel operators. 1. For unusual or risky operations that are anticipated to be encountered, the CFR does require additional information to be provided to the crew. Vessels that are in special service, such as towing, lifting, and fishing face specific stability risks not faced by other vessels and it is expected they would have additional guidance on board relative to those risks. 2. Extreme condition guidance could be included as information on measures the vessel could take to reduce risks of loss of stability and maintenance of watertight integrity. 3. While the stability booklet would still take precedence, the use of stability software could be encouraged for smaller vessels for which loading programs (or stability instruments) are not required. Loading programs could be an approved type (class or USCG) or reviewed and approved by a qualified individual (as defined by 46 CFR Subchapter C § 28.510). Larger Ocean-Going Vessels For larger ocean-going vessels with international load line and SOLAS certificates, stability information is provided by a stability booklet and damage stability information is provided by a damage stability booklet and plans, per SOLAS, Chapter II-1, Part B-4, Regulation 19, Damage control information. However, the stability booklet is based on intact static stability requirements and generally does not provide the vessel much guidance on stability issues in extreme weather conditions, when vessel stability could be at risk. Additional information on extreme weather stability risks could be of use to vessel operators who may encounter those conditions. The following are examples of the type of information that might be useful to vessel operators at sea.

67 Further research on this topic is needed to determine if the provision of this type of information is appropriate on board. 1. Providing extreme condition guidance information to the vessel operator as a supplement to the stability document could be useful. If added to the stability instrument, information indicating the effects of wind speeds up to 100 knots, for example, on vessel heel at all apparent wind angles (such as at every 15 degrees) for a range of drafts, vessel cargo profiles, and GM could be beneficial. 2. Information on the susceptibility of the vessel to parametric rolls, broaching, and stability loss in quartering seas would be useful to include in the extreme condition guidance. As dynamic stability analytical processes and requirements are added to vessel stability rules and regulations (e.g. dynamic stability requirements under consideration for the IS Code), the results of these calculations on stability risks to vessels in extreme conditions would be added to the guidance information provided to the vessel operator. Such information would be encouraged to be provided for older vessels, even if the vessel were not required to comply with the updated stability code.21 3. Extreme condition guidance could include information on measures the vessel could take to reduce risks of loss of stability and maintenance of watertight integrity. Information and checklists would be provided on all openings with covers that should be closed or maintained closed while at sea, and, in particular, those that should be closed in severe wind and wave conditions. Periodic training and drills would be 21 See MSC.1/Circ.1228, Revised Guidance to the master for avoiding dangerous situations in adverse weather and sea conditions, January 2007.

68 carried out to ensure crew knowledge of where the closures are located and the circumstances in which they should be closed and to ensure fast and efficient action when the need for closure arises. 4. Limits on propulsion system operation at extreme heel would be checked and guidance provided on the potential risks with the loss of propulsion. 5. Risks of cargo shifting or breaking loose would be described for standard cargo types that are carried on the vessel. 6. Older vessels would be required to have on board damage stability information to meet current SOLAS standards. The El Faro Investigation During the hearings by the Marine Board of Investigation into the loss of the El Faro, there was interest by the board as to whether adequate information was available to the operating crew about the stability risks from the severe weather encountered and the impact of the flooding found in one or more holds. In particular, the board’s Safety Recommendation Number 2 questions whether there was adequate information about the need to close the fire and weather dampers. Safety Recommendation Number 16 recommends that damage control information be available on all existing cargo vessels. Safety Recommendation Number 17 recommends that damage control training and drill requirements be implemented for commercial, inspected vessels. The USCG Commandant indicates agreement (or partial agreement) with these safety recommendations in the Final Action Memo to the Marine Board Report.22 22 The USCG Final Action Memo is available at https://media.defense.gov/2017/Dec/21/2001859858/-1/- 1/0/EL%20FARO%20FINAL%20ACTION%20MEMO.PDF.

69 References Anand, S., N. Keren, M. J. Tretter, Y. Wang, T. M. O’Connor, and M. S. Mannan. Harnessing data mining to explore incident databases. Journal of Hazardous Materials, 130(1–2), 2006, pp. 33–41. Blanco-Novoa, O., T. M. Fernandez-Carames, P. Fraga-Lamas, and M. A. Vilar-Montesinos. A Practical Evaluation of Commercial Industrial Augmented Reality Systems in an Industry 4.0 Shipyard. IEEE Access, Vol. 6, 2018, pp. 8021–8218. Chu, X., I. Ilyas, and P. Papotti. Holistic data cleaning: Putting violations into context. Institute of Electronics and Electrical Engineers, 29th International Conference on Data Engineering, 2013, pp. 458–469. Díaz-Casas, V., A. M. Doce, P. T. Martinez, S. F. Gonzalez, and M. Vilar. Shipyards of the 21st century: Industrial internet of things on site. In Progress in Maritime Technology and Engineering: Proceedings of the 4th International Conference on Maritime Technology and Engineering (MARTECH 2018). CRC Press, 2018, p. 439. Ellis, J. Analysis of accidents and incidents occurring during transport of packaged dangerous goods by sea. Safety Science, Vol. 49:8–9, 2011, pp. 1231–1237. European Maritime Safety Agency, 2018. http://www.emsa.europa.eu, retrieved April 24, 2018. Federal Bureau of Maritime Casualty Investigation (Germany), 2018. http://www.bsu- bund.de/EN, retrieved April 24, 2018. Grabowski, M., J. R. Merrick, J. Harrald, T. Mazzuchi, and J. Van Dorp. Risk modeling in distributed, large-scale systems. IEEE Transactions on Systems, Man, and Cybernetics—Part A: Systems and Humans, Vol. 30:6, 2000, pp. 651–660. Grabowski, M. R., C. A. Dorsey, and B. Wang. Data Analytics to Enhance Prevention Risk- Based Allocation of Vessel Safety and Environmental Compliance Inspections and Examinations. U.S. Coast Guard, Marine Safety Center, Washington, DC, 2018. Hassel, M., B. E. Asbjørnslett, and L. P. Hole. Underreporting of maritime accidents to vessel accident databases. Accident Analysis & Prevention, Vol. 43(6), 2011, pp. 2053–2063. Hellenic Marine Accident and Event Investigative Service, 2018. http://www.hbmci.gov.gr, retrieved April 24, 2018. International Maritime Organization. Global Integrated Shipping Information System. 2018. http://gisis.imo.org/Public/MCI/Default.aspx, retrieved April 24, 2018. International Maritime Organization. Report to the Maritime Safety Committee. 5/15, Section 6, February 15, 2018.

70 Kalyvas, C., A. Kokkos, and T. Tzouramanis. A survey of official online sources of high-quality free-of-charge geospatial data for maritime geographic information systems applications. Information Systems, Vol. 65, 2017, pp. 36–51. Khaleghi, B., A. Khamis, F. Karray, and S. Razavi. Multisensor data fusion: A review of the state of the art. Information Fusion, Vol. 14:1, 2013, pp. 28–44. Marine Accident Investigation Board. Accident Reports. 2018, https://www.gov.uk/maib-reports, retrieved April 24, 2018. Mazaheri, A., J. Montewka, J. Nisula, and P. Kujala. Usability of accident and incident reports for evidence-based risk modeling–A case study on ship grounding reports. Safety Science, Vol. 76, 2015, pp. 202–214. Mullai, A., and U. Paulsson. A grounded theory model for analysis of marine accidents. Accident Analysis & Prevention, Vol. 43(4), 2011, pp. 1590–1603. National Research Council. Ship Collisions with Bridges: The Nature of the Accidents, Their Prevention and Mitigation. National Academy Press, Washington, DC, 1983. National Research Council. Crew Size and Maritime Safety. National Academy Press, Washington, DC, 1990. National Research Council. Fishing Vessel Safety. National Academy Press, Washington, DC, 1991. National Research Council. Minding the Helm: Marine Navigation and Piloting. National Academy Press, Washington, DC, 1994. National Research Council. Applying Advanced Information Systems to Ports and Waterways Management. National Academy Press, Washington, DC, 1999. National Research Council. Special Report 273: Shipboard Automatic Identification System Displays: Meeting the Needs of Mariners. National Academy Press, Washington, DC, 2003. National Transportation Safety Board. Transportation Safety Databases. Washington, DC, 2002. National Transportation Safety Board. Marine Accidents Report. 2018, http://www.ntsb.gov/investigations/AccidentReports/Pages/marine.aspx, retrieved April 24, 2018. Storgård, J., I. Erdogan, J. Lappalainen, and U. Tapaninen. 2012. Developing Incident and Near Miss Reporting in the Maritime Industry—a Case Study on the Baltic Sea. Procedia: Social and Behavioral Sciences, Vol. 48, 2012, pp. 1010–1021.

71 U.S. Committee on the Marine Transportation System. U.S. Arctic Marine Transportation System: Overview and Priorities for Action. Washington, DC, 2013. U.S. Department of Homeland Security. National Incident Management System. Washington, DC, 2008.