On a Course to Automation

Captain Eero Lehtovaara
Captain Eero Lehtovaara
ABB Marine and Ports

16 August 2018


ABB has developed a conceptual framework and identified the technological requirements for switching to an automated navigational watch during the ocean leg of long-haul international voyages. This article is a summary of a white paper prepared by the author and Kaveli Tervo, Global Program Manager of ABB Marine & Ports.

Safer navigation with nobody watching

The number of crew working on deep-sea ships has followed a downward trend for decades, but still today an officer of the watch (OOW) is required on the bridge regardless of conditions. This is justified when a vessel is approaching other ships or sailing in areas where traffic is anticipated but seems unwarranted during uneventful ocean crossings.

In good and clear conditions, the OOW can work an entire shift without touching any equipment or doing anything but looking at radar screens and out of the window to make sure that nothing is amiss. Nevertheless, the OOW is responsible for complying with COLREGs, checking navigational equipment in use at regular intervals, preparing, executing and monitoring a safe passage plan, requesting assistance when necessary, alerting the master when the need arises, and not leaving the bridge unattended. He/she must also verify the vessel’s position to form a ‘mental picture’ of its situation. Even in the 21st Century, operating and regularly checking the ship’s compasses and depth sounder are fundamental requirements - particularly in poorly charted areas.

In themselves, these responsibilities can cause mental fatigue and loss of alertness, which might reasonably be expected to result in a delayed or impaired response when an event requiring manual intervention does occur.

Recent advances in sensor technology, computing power and data analytics have broadened the scope of what can be accomplished with automation, sparking a renewed interest in the notion of unmanned and/or autonomous vessels. Before such remote or self-controlled vessels take sail, there is no reason the same technologies could not be applied to reduce the cognitive burden on today’s OOW.

This realisation has prompted ABB to explore the idea of the conditionally and periodically unmanned bridge, which would reduce fatigue, enable more efficient utilization of vessel crew, and, above all, increase safety.

Regulatory needs

STCW lays out what crew are needed on the bridge in different conditions: Bridge status 1 (B1) – day time, good conditions – only requires an OOW; at night in good conditions (B2) he must be accompanied by a lookout; in special conditions (B3), an additional helmsman must also be present. Reflecting this terminology, ABB codenamed its unmanned watch solution ‘B0’, with a nod to the ‘E0’ status accepted as denoting unmanned machinery spaces or engine control room. The OOW also bears legal responsibility in the command hierarchy when the Unmanned Engine Control Room status is lost: he/she must decide when it is safe to revert to E0. If the bridge is unmanned, an alarm causing the loss of E0 status would summon an engineer, but also call the OOW to the bridge, so cancelling B0 status.

Technology requirements/conditions

From a technological perspective, the system would have to fulfil the role of the OOW in monitoring the status of the navigation equipment, to identify gyrocompass errors, and oversee radar performance or the stability of the GNSS. Cameras and possibly other sensors which offer equivalent fields of view and range as a human eye will assume the role of the OOW looking out the window, which is a regulatory requirement to prevent over-reliance on radar.

SOLAS sets out conditions for visibility from the bridge. From the main conning position, the vessel must have 225 degrees visibility with a view of the sea surface not ‘obscured by more than two ship lengths, or 500m, whichever is the less, forward of the bow to 10° on either side under all conditions of draught, trim and deck cargo.’ When designing an unmanned bridge, these requirements would need to be met by a virtual line of sight.

To provide equivalency for the SOLAS Ch. 5, req. 2, there needs to be automated target identification and classification. ABB is already deploying solutions for carrying out this task as an aid to bridge staff. To ensure sufficient resilience, ABB requires a minimum of two independent data sources for all measurement and situational awareness technologies. The extra layers of protection are especially beneficial during night time operation where, for instance, a nearby vessel might be identified by either infra-red camera or by a conventional daytime camera detecting its lights.

ABB envisages that the OOW would engage B0 or unmanned watch mode when the weather and visibility are good, with no fixed objects visible in the forward sector nor any vessels visible with a closest point of approach (CPA) or time for closest point of approach (TCPA) below a predetermined threshold: in short, when their role is to ensure nothing is done.

The critical task is setting appropriate thresholds for CPA and TCPA. In general, radars have a range of 96 nautical miles. A target at this range will take 6h24m to reach if the vessel is sailing at 15 knots. However, if the other vessel is also sailing at 15 knots, the vessels would meet in half that time. The thresholds could therefore be either absolute or fixed values or be set dynamically according to the speeds of the vessels and allowable reaction time. A procedure or checklist to ensure that all systems are operating normally would have to be followed to ensure safe handover.

The system must also recognise and continuously monitor land. If a shoreline is detected on the horizon in the forward sector, the OOW is summoned and B0 status terminated. Here, visual systems will be supported by microphones and supporting software capable of recognising unexpected audio. Additionally machine learning could be employed to manage situations where no radio watch is kept, to send and alert the OOW to critical messages via a mobile device. The OOW should have immediate access to the navigation, radar, camera and other relevant sensor data during B0 status, using a similar method as the E0 or UMS systems which inform Engineers today that the alarm has been triggered. This too could be performed via a mobile device.

Conclusions

Taking all this into account, the unmanned watch would appear to be acceptable so long as the technical solution demonstrates equivalency with or better than human capabilities to meet current regulatory requirements.

ABB understands that the idea of an unmanned bridge – even for limited periods in benign conditions – is prickly from both regulatory and cultural perspectives. However, the industry is reaching a point where (according to BIMCO) the pool of competent seafarers is shrinking while the safety of our crews, cargo and ships remains as important as ever. It is ABB’s intention to leverage technology, which is advancing and maturing rapidly, to work alongside and reduce the burden on competent ship’s officers and thereby improve their decision making.

Enabling a better quality of rest when conditions are good and there is no need for major navigation actions will reduce fatigue and have a net positive impact on crew alertness when approaching coastal areas or routes which do have traffic. Moreover, working hours during the ocean crossing part of the voyage could be closer to normal office hours. Meanwhile, raising the alarm if something unusual does happen during the ‘quiet’ part of the voyage, the bridge team will respond more adeptly than if the situation had built up slowly over time in front of a tired OOW working alone.•