Due to the increasing seaborne trade, shortage of officers, constant work of crew on machinery during voyage and the large contribution of human error to shipping accidents it is time for shipbuilding to take the next step in automation. Unmanned shipping is the way to move forward and since the engine room houses some of the most important machinery in the ship, this is the equipment which has to be made more reliable first. Improving the reliability of ship machinery is the goal of this thesis, and this will be done for ship types with a simple design and relatively simple equipment, namely general cargo ships, container ships, bulk carriers and oil and chemical tankers. In order to increase the reliability of the machinery, it will first have to be determined which failures occur to the machinery. The main engine, steering gear, fuel system, electrical system, cooling water system, diesel generator, shafting and other have all been found to be responsible for fatal technical failures in the past. Since these first five systems are responsible for the largest part of the failures, the focus of this thesis will be on these systems. These systems are crucial for the actual sailing of the ship, so making these systems more reliable makes great effort regarding unmanned sailing. With the failures known, solutions have been developed for each individual failure. These solutions should have an equal or higher reliability compared to the initial solution. For the found solutions it can then be determined what their reliability improvement is. Determining the reliability improvement is done by knowing the failure rate of the machinery, which has also been determined. A list can then be made to determine which solutions provide the most reliability improvement. The cost of the best solutions is calculated next. In order for the unmanned ship to sail the oceans in the near future, it has to be financially competitive to the manned ship. When looking at the causes of the failures it was found that human error and improper maintenance are both responsible for almost 20% of the failures. Performing better maintenance was though not found to realize much reliability improvement, just as an alternative solution was not found to realize much reliability improvement. Solving the maintenance related failures or choosing an alternative system is thus not the best solution to the failures. The best type of solution was found to be a type of redundancy. Solutions to the main engine failures, steering gear failures and fuel system failures were found to achieve the most reliability improvement. This was to be expected since these systems are also responsible for the highest number of failures. When making a ship unmanned, equipment such as accommodation, fresh water systems, HVAC and others do not have to be installed. Money that does not have to be spend on this equipment is the budget for the reliability solutions. The cost of the three solutions with the most reliability improvement was found to fit perfectly within the budget that is available. It can thus be concluded that it is possible to increase the reliability of ship machinery while staying financially competitive to the manned ship. Budget is then also still available to pay for other systems that are needed for unmanned shipping. Redundancy was found as being the best type of solution. If all the best solutions are applied, the machinery would have a mean time between failures of nine years. Finding solutions to the diesel generator failures, shafting failures and other failures is needed to increase the reliability even further. In order to improve the reliability of the machinery even further, more detailed failure data has to be known. Of half of the failures there is currently not enough information. Better knowledge of the failure distribution will help to determine better solutions. It also helps to know to what ship types these failures occurred to. Failures in redundant systems are not within the failure data. Having this data really helps to get a better view of the problem. Knowing what failures the crew prevents during operation and knowing more precisely which failures they induce also helps to get more detailed knowledge of the machinery failures. If better failure data is available, a more detailed failure rate can be determined. When it also known which level of reliability is required, the more detailed failure rate can be used to determine whether this level of reliability can be reached. The here named advises would though all require extra research and are therefore recommendations for future work.