Service scheduling
Description / objective
The Service Scheduling function operationalizes the Network plans. Scheduling is normally performed at individual route level, although in some cases the scheduling is performed for a number of routes where there are overlaps or potential for resource optimisation.
Scheduling is a key competence of a bus operator, as it determines the resources required to perform a specified service plan. This is especially important in competitive tendering situations, where an operator with excellent scheduling capability can invariably find efficiencies and creative solutions, and outbid a competitor with weaker scheduling skills. A highly competitive Operator is also likely to change labour agreements (subject to compliance with labour law) to remove constraints from efficient and flexible scheduling.
The specifications for a route can emerge from a formal network planning study – this is particularly the case where new routes are considered, or where all the routes in an area/corridor are being reviewed. Alternatively, modifications to an existing route are required - e.g. due to the need for more buses on a growing route, or due to falling ridership, or due to difficulties to achieve the schedule due to traffic disruptions.
In any case, the basic parameters for the route are defined – route travelled, terminal points, stops served, first/last trip times, desired service levels throughput the day. These may be ‘absolute’ requirements, or they may be an initial specification that could be adapted as a result of the scheduling exercise. Additional inputs are the travel times for the route (sometimes by route section) by direction and time of day.
There are five key tasks associated with the Service Scheduling function. They may all be included in the Service Scheduling Function, or some may be performed separately:
- Generation of the detailed timetable
- Vehicle scheduling, so that all trips are covered and assigned to a specific vehicle. This is performed in a way that optimises the number of vehicles required. It also takes account of when vehicles enter and leave service (e.g. during the off-peak) and where they vehicles are stored at this time
- Driver scheduling, so that all pieces of vehicle work are covered and assigned to a specific piece of driver work. This is performed in a way that optimises the number of drivers required (and associated conductors, where relevant). This must also take account of driver hours, meal breaks, rules regarding maximum work periods (part-shift, total shift, weekly total), and of local labour agreements. In some cases, it must also take account of the time required for the driver to get to a start/finish point when travelling without the vehicle
- Resource requirement calculation – number of buses, drivers, kms operated, by time of day and by day type
- Generation of ‘running boards’ – i.e. the detailed piece of work associated with each vehicle duty and each driver duty
Some systems include a cost function, so that the resource requirements (vehicles, personnel, fuel) are calculated along with the associated cost. This allows the scheduler to take decisions on whether the proposed solution is viable or optimal – especially if ridership and revenue estimates can also be calculated. The scheduler may decide to adjust the input parameters (e.g. adjust the service intervals at certain times of the day, adjust first or last trip time, change the vehicle mix) to achieve a better financial outcome. Of course, any adjustments to the basic parameters need to be discussed with the other stakeholders.
Technologies, data and resources
Until the 1980’s, all scheduling was performed manually. Computerised scheduling emerged during the 1980’s, but was rather cumbersome in terms of inputting and compiling, and it could take many hours to perform a single run of the program. As computational power of PCs increased exponentially, these limitations disappeared, and computerised scheduling has become a common tool for bus operators and transport authorities.
The Service Scheduling function is usually integrated with other ITS and IT systems. It has a close linkage with three main systems:
- AVM system: The scheduling software provides the AVM system with the primary base data about the planned service, departure times, intermediate times, vehicle and driver duties, meal breaks, etc. In turn, the AVM system can export route and sectional journey times, and their variations (by direction, by time of day, by day type) as direct input to the scheduling process.
- Rostering/assignment system: The scheduling system exports the vehicle and driver duties as base data to the rostering system, which performs the daily assignment of drivers and vehicles.
- Traveller Information systems: The scheduling system exports the timetables to the various traveller information systems.
Simple scheduling packages are little more than spreadsheets, which arrange the information that was previously done in pencil by a manual scheduler, and performs supporting computations. This can often be sufficient for small companies or smaller routes.
By contrast, more advanced scheduling packages are ‘resource-hungry’. They require significant amounts of input data (which may be provided by the AVM system or traffic control system), and may use advanced algorithms to optimise timetables, vehicle and driver resources, and cost.
Advantages and cautions
The primary advantages of ITS-supported Service Scheduling are:
- Extensive and advanced computations and optimisation can be performed, which were previously beyond all but the most expert schedulers
- Journey time data can be transferred automatically from the AVM system. This allows greater precision of the schedule, saves on data input costs, and eliminates human errors
- Schedules can be generated more quickly. This allows the Operator to be more responsive to changes in demand or operating conditions. It also allows Operator to look at a number of options, where previously the scheduling effort involved would have been prohibitive.
- Automatic generation of running boards and direct inputs to the AVM and rostering systems
- Competitive advantage in tendering situations
- Quicker and less expensive to develop a competent schedulers
The primary cautions are:
- Since it is automated, users of scheduling software may not develop the skills to ‘tweak’ the solution offered by the software, thus either gaining a bit more efficiency, or making a marginal solution compliant with timetable specification or labour agreements
Case Studies
Dublin, Zurich