At 11:46 AM on 19 July, 2005 Savannah Express (SE), made heavy contact with a link span in Southampton. SE was one of the most technically advanced ships in service. The operation of the main engine was done by a highly automated, computer controlled electro-hydraulic system. This system had suffered a number of technical problems since SE had left the builder’s yard. At the previous port of call, a service engineer had attended the vessel to rectify various guarantee claims. SE left without certain spare parts associated to the electro-hydraulic system and a faulty hydraulic pump indicating the pump “stuck” in the 100% ahead position.

At 7:00 AM SE approaches Southampton and at 7:29AM there was a shutdown of the main engine. Various hydraulic pressure alarms occur simultaneously and the crew works in parallel processes to solve the problem. The crew turned the main engine successfully on air and was under the impression that the fault was corrected .Upon visual inspection, the both control units looked identical but turned out not to control the same equipment. As SE approached Southampton she got a shut down alarm and slowly turned starboard she made heavy contact with the pontoons of the link span, severely damaging it.

The purpose of this paper is to elaborate on the complexity of automated systems and lack of transparency within those systems. In this specific case it led the crew to take actions which intuitively felt right, but ended being wrong. This paper will examine how the alarm system hinders fault finding undertakings and identifies how proper decision support systems could have helped the crew to understand the emerging events which led to the final accident.