2.6 Congruence of Navigation Data

　

 About 20 years ago when no GPS was available to the public, the position of a ship was treated as a point and the position fix was inaccurate. Everybody was aware of this fact.

　

 Today, accurate electronic chart material is available and ship's positions can be measured down to an accuracy of less than a single metre. Ships are about 300 and more metres long. So the question of the ship's position has acquired a new aspect. Which location on the ship is the reference that is used for the definition of the ship's position?

　

 Unfortunately there is no generally accepted regulation for the definition of a ship's position. It can be defined anywhere on the ship.

　

 The IEC has proposed the management of ship's data and references as follows:

 'The different locations of antennas and other sensor receiving units should be transformed to a common reference location on the ship. This location should be predefined by the manufacturer or adaptable to the special conditions of each ship. In order to align different location information to a common reference location information, the consistent common reference system should be established. All functions should use the identical values for reference parameters of the same type.'[3]

　

 This means that manufacturers of navigation systems should (and do) define a common reference location in the way that best suits the navigation system's strategy of internal data processing. This reference location is essential for accurate processing of track calculation, target tracking, CPA/TCPA calculation, collision avoidance, etc.

　

 In some Integrated Navigation Systems (INS), all valid position sensor data are weighted and filtered to compute the most certain position information out of all available navigation sensors.

 Other Integrated Bridge Systems (IBS) use the best available or manually selectable sensor for further processing and use the secondary or other sensors as backup systems.

　

 Hence the accuracy with which a ship's position can be determined is not only defined by the position sensor and its accuracy, but also depends on the data processing strategy and the definition of the reference location on the ship itself.

　

 This information is essential for the mariner and he has to be aware of if.

　

2.7 GPS Antenna Location

　

 Besides the location of the GPS antenna also the height of the antenna should also be taken into account for the sensor simulation. The GPS antennas are mounted on the ship's superstructure at heights of about 35 metres above sea level to guarantee an unobstructed view of satellites. When the ship is making a turn, the heel of the ship will move the GPS antenna at high angular speed in the opposite direction. This is interpreted by the GPS navigation system as an apparent movement of the ship in the direction opposite to that of the turn, which results in wrong speed information and influences the course made good. The opposite effect is generated at the end of the turn.

　

 Additionally, the position of the ship will be offset depending on the heel. This means that a GPS navigation system with an antenna mounted 30 metres above sea level will display a position offset of about 1 metre per 2 degrees of heel.

　

 Both effects influence the GPS navigation sensor accuracy in rough seas.

　

2.8 Speed Log Sensor Location

　

 Today, dual axis Doppler speed logs are installed on most ships. The Doppler speed log provides bottom and water track speed information in the ship's longitudinal and transversal directions at the transducer location.

　

 The measured speed log data are valid at the location of the speed log sensor only. Before these values are processed, at least the transversal speed has to be transformed to the respective sensor location, e.g. the radar antenna.

　

 Neglecting these facts for ship design, sensor configuration or simulator design may result in strange behaviours of tracked targets, CPA/TCPA calculations, track-keeping systems and during manoeuvring and docking procedures, etc.

　

 The mariner should also be aware of the connection between the speed and the location on the ship for which the speed display is valid. Depending on the implementation of the navigation system, different speed values may be displayed on the speed log display, the radar display and other navigation equipment.

　

 Figure 4 may convey an idea of consequences on ARPA information if the measured transversal speed of the bow speed log sensor is not correctly transformed to the radar antenna position. The figure shows a turning ownship, where the bow is turning to starboard while the stern of the ship is turning to port. Wrong speed information given to the ARPA system would thus result in wrong ARPA information.

　

　

2.9 Gyro Sensor Location

　

 Another vital item is the gyro compass if used to shoot visual bearings which are taken in relation to radar and ECDIS display information.

　

 The gyro compass is located in the wheelhouse and hence all visual bearings are taken from there. While the radar system references all electronic bearings relative to the radar antenna position, there may be a difference for bearings of objects close to the ownship, e.g. during berthing or close quarter manoeuvres. This error applies especially when the radar antenna is situated far away from the gyro compass location which is used to shoot the visual bearing.

　

2.10 ECDIS

　

 The ECDIS chart too has to be treated as a kind of navigation sensor which may be afflicted with errors. The chart information may be wrongly referenced, offset, omitted or false. Overlaying of radar, tracks, routes or bearing information from other navigation equipment on the ECDIS may confuse the mariner if the data are not in line with the ECDIS chart.

　

 In particular, offset errors between the ECDIS chart and overlay of radar echoes are frequently observed. Assuming a correct ECDIS chart, a wrongly selected geodetic datum or a wrong position sensor configuration or setting may be the reason for this.

　

2.11 Automatic Identification Systems

　

 In July 2002, the Automatic Identification System (AIS) became obligatory for all new ships. The aspects of processing and display of navigation data received by AIS should also be taken into consideration for sensor simulation. Furthermore the navigation data received may be afflicted by sensor effects and errors which have not been detected by the transmitting ship's navigation system.

　

 Comparing received AIS data with the navigation data provided by the ownship's navigation system may require the mariner to decide which information is right and which is wrong.

　

 A simulator incorporating advanced sensor simulation taking the described effects into account can be used to demonstrate realistically the effects of sensor behaviour and malfunctions.

　

 However without adequate representation of the differences between reality and the sensor display of reality there will be little training effect for the mariner or student.

　

 While the student always acquires the sensor data displayed on his navigation systems, the instructor has to have the possibility of superimposing these sensor data as seen by the student on the 'real world' situation as displayed on the ECDIS of the instructor station. This would make the differences between the sensor and system data visible for the instructor.

　

 Figure 5 proposes a possible way of achieving such implementation. The figure shows the real system track together with the superimposed sensor track of the ownship.

 Along with the graphical situation, the sensor position, heading, speed made good and cross track error, relative to the system track should be presented in numerical format.

　

　

 An additional graphical presentation should be made available to the instructor to display the transversal speed components along the ship's hull together with the dynamical location of the ship's pivot point. This allows the presentation of the relationship between the turn rate of the ship, the transversal speed components and the ship's pivot point.

　

 All presentations on the instructor station situation display should be available during exercises and also during debriefing and replay. At any time the instructor should be able to select any of the available sensor data to be superimposed on his situation display.

　

 With such tools, sensor effects and malfunctions could be demonstrated to the students. For further assessment, the deviation values should also be shown in a parameter plot or be available in tabular format.

　

 The selection of sensor malfunctions and effects should be controlled by the instructor. The following figures present some sample layouts of windows for selection of sensor effects and malfunctions.

　

　

　

　

　

 Although only some of the navigation sensors and effects are described here, the consequences of inaccurate or wrong position, speed and heading or referenced sensor locations can be seen. The entire chain of sensor data processing is very complex and sensitive to any type of dynamic disturbances and also to failures in system configurations.

　

 All manufacturers of navigation systems are aware of the facts described and many precautions and safety functions are built into real systems to compensate effects and to prevent malfunctions.

　

 However there is still a possibility of error which cannot be detected by the navigation system, and that could lead to extremely dangerous situations.

　

 A couple of years ago, seafarers had limited navigation equipment available to them, resulting in uncertain position fixes. However the seafarer was a link in the chain of data processing and was thus well aware of the limitations of his navigation sensors, and this resulted in prudent seamanship.

　

 Today the seafarer simply has to monitor the ship's navigation systems or Integrated Bridge Systems presenting position data with an accuracy down to a metre. With the increasing number and complexity of individual navigation systems on board ships, the safety of the ship becomes more and more reliant on technology and on the capability of the mariner to interpret the displayed information correctly.

　

 The accuracy of navigation sensors is very high, which results in a good representation of the real world. So the gap between reality and the human perception of reality is very small, but seems to be worth simulating, because:

　

　

 This lecture has given an introduction to a possible way of simulating the gap between reality and the human perception of reality by implementing realistic sensor simulation in ship handling or navigation simulators.

 It can be used as a basis for further definition of requirements and needs concerning the simulation of sensor effects and malfunctions.

　

 I am pleased to acknowledge the support given by Captain Herman J. von Morgenstern and Holger Stoltenberg of the Institute of Ship Operation, Sea Transport and Simulation 'ISSUS' in Hamburg, Germany; their assistance and the fruitful discussions in which they took part have been very helpful in preparation of this lecture.

 I would also like to acknowledge the help given by Mr. Hubert Hayek and Mr. Wilfried Frerichs of SAM Electronics who provided detailed information on Integrated Bridge Systems.

　

[1] IMO, "Performance Standards for Electronic Chart Display and Information Systems (ECDIS)", International Maritime Organization Resolution A.817(19), 23 November 1995

[2] "MALAYSIA SHIPPING NOTICE NPM 11/1999"

http: //www.marine.gov.my/npsaudagar1199.html &

[3] IEC:2000 International Electrotechnical Commission, Maritime Navigation and Radiocommunication Equipment and Systems "Track Control Systems" - DRAFT -

[4] ATLAS NACOS xx-3 Navigation and Command System - Alarm Management ED 3030 G 672 Ed. 01 (2001-11 / 80)

[5] ATLAS NACOS xx-3 Navigation and Command System - System Description ED 3030 G 562 Ed. 01 (2002-01 / 78)

[6] Rothblum, Sanquist, Lee, and McCallum. "Identifying the Effects of Shipboard Automation on Mariner Qualifications and Training and Equipment Design" ISHFOB '95 Intern. Symposium - Human Factor on Board. Bremen Nov. 15-17, 1995

　

 Mr. Martin Staden is working for STN ATLAS Elektronik Bremen, Germany since 1980. He holds a position as a Senior Technical Manager for large scale Maritime Simulator Projects. As the Leader of the Maritime Simulation Development Group he was responsible for the design and development of the ANS4000 / ANS5000 Maritime Simulator Family used for Ship Handling-, Radar-, ECDIS-, Vessel Traffic Services-, Engine Room-, and various naval application based simulators. Previous experience includes design and development of simulators for naval surface and subsurface platforms including simulation of modern radar, sonar, weapon and electronic warfare systems. He holds a Master of Science in Telecommunication and Information Technology acquired at the "Fachhochschule für Technik" in Osnabrueck, Germany.