Maintenance Manager

Maintenance planning is usually performed manually in the railway sector. Planners need to schedule a large number of activities by considering multiple constraints (related to the availability of personnel, inventory, and specialized equipment). Moreover, maintenance activities must be synchronized with the regular operations of the rolling stock. Such a planning process is very complex. Creating even a feasible plan, by hand, is time consuming and ineffective. Sub-optimal scheduling decisions not only increase the maintenance cost substantially, but reduce the time in which rolling-stock units are operative. The focus of this project has been on developing new optimization algorithms that can support the planning process. With them, planners can get suggestions on when to schedule each maintenance activity, such that resources at the workshop are utilized as efficiently as possible. Activities can be suggested to be executed earlier than required if it allows the workshop to streamline its operations. This allows not only to decrease workload peaks and idle times, but to increase the throughput of rolling stock. This project, supported by the research performed by SINTEF, has allowed us to prototype an optimization-based tool for maintenance planning. This tool will help us, and our customers, to reduce the maintenance costs and provide a better and more reliable service. These efforts, altogether, will provide us with a significant advantage over our competitors. The prototype has been tested and improved in the final period of the project. The planners using the tool have lauded the user interface for its user friendliness and ability to support their work flow. They found that the tool produces good maintenance schedules requiring much less effort from them than before.

The most significant outcome of MainMan is the prototype of a novel tool for scheduling the arrivals and operations of trains at a maintenance depot-workshop. Once implemented in an industrial tool, it will lead to several improvements both in maintenance processes and for the railway system and the society. Maintenance Process: 1. The planners will increase their efficiency and have an estimated reduction of current workload of 75%. 2. Re-planning becomes feasible, i.e. last day adjustments due to unexpected changes 3. The logistic department will be able to better supply components and identify lack of parts at early stages 4. The production increase can be estimated in the range 10% to 20% Railway System and society: More punctual and precise maintenance scheduling will improve 1. system punctuality, with benefits for passengers and society 2. usage of rolling stock, reducing the number of circulating trains and so global costs and impact on the environment

Currently, maintenance planning in the railroad domain is predominantly done manually and involves: Crew scheduling (assigning maintenance tasks to crews dependent on skill set) and job shop scheduling (assign time slots for rail vehicles to certain depot workstations). These activities should be synchronised and coordinated with rail vehicles revenue generating activity (where the vehicles are scheduled to transporting goods or passengers). Individually, these three planning tasks are very complex. Finding a manual solution is time consuming and highly inefficient. Due to the substantial fixed costs involved, inefficient solutions are very costly. Few tools exist to assist in these planning task, they normally focus on myopic parts of the overall problem and solve the planning problems independently and then merge them-thereby losing the global perspective needed to get the efficiency level needed. Individually, these different planning problems have been well studied in the optimisation literature and other industries optimisation based planning tools frequently. E.g. optimisation tools for airplane maintenance has been successfully used for decades. It is time to learn from other industries and apply optimisation techniques to increase the efficiency in railroad maintenance. MainMan will give us the research from SINTEF Optimisation, to create an optimisation based maintenance planning engine, that can solve these interconnected planning problems simultaneously while respecting the shared and limited resources. To us and our customers in the project (NSB, Flytoget and CargoNet), this will reduced costs and lead to a better and more reliable service. Partly due to more efficient crew and maintenance task planning. By considering railroad vehicles seasonal peaks we might reduced the total number of vehicles our customers needs to carry out their commitments. Altogether these efforts will give us a significant competitive advantage towards our competitors.