How Much Should You Depend on Drag-and-Drop Operations on Gantt Chart to Get Good Schedule?
Prasad Velaga, PhD
Scheduling Specialist for High-Variety, Order-Driven Production and Resource-Constrained Projects
Drag-and-drop functionality on Gantt chart is seen by many scheduling personnel as an important feature of software for scheduling production/project operations. Many scheduling software sellers proudly talk to potential customers about drag-and-drop feature of their software.
In the past, schedulers used to manually develop production schedules on whiteboards in the form of Gantt charts. They found it cumbersome, laborious and time consuming to manually update the schedule on the physical boards in response to significant changes in production. This effort often involved rescheduling of operations over timeline and/or reassignment of resources to operations. The representation of operations by magnetic pieces on the board reduced the effort to some extent. The availability of powerful personal computers with graphic card not only enabled quick generation of schedules and display of the corresponding Gantt charts on electronic monitor but also facilitated manual changes in a schedule by the dragging and dropping (moving around) of operations on Gantt chart. People using physical whiteboards found that the drag-and-drop feature on Gantt chart can greatly reduce the effort involved in manual changes in the schedule on Gantt chart. Production supervisors who make real-time decisions for assigning machines or workers to operations tend to like drag-and-drop functionality. A drag-and-drop operation is usually performed for rescheduling of an operation and/or for changing resource assignment to operations over a short time interval.
A scheduler may see a need to perform a drag-drop operation when he finds some part of the schedule infeasible, irrational or improvable. A software-generated schedule can be infeasible, irrational or improvable when the scheduling logic used by the software is very weak. Drag-and-drop feature is mostly used for correcting a schedule or improving schedule quality over short time intervals. If a machine is overloaded while keeping a similar machine idle in the schedule, some jobs can be moved from busy machine to idle machine on Gantt chart by drag-and-drop operations. A drag-and-drop operation may however have an invisible ripple effect on the entire schedule with a risk of violating some constraints and operation precedence relations and causing resource overloading over schedule horizon.
The drag-and-drop feature looks elegant for people who focus on schedules of short intervals. Some people feel that drag-and-drop functionality gives them more control on creation and revision of schedule. However, vendors provide this functionality not only for this reason but also to compensate the weakness of their hidden scheduling logic. If users see a big, undesirable necessity to depend on this functionality, then it is a sign of major weakness of scheduling logic in the software.
No scheduling software uses perfect logic to be able to generate the best schedules in every target industry. Therefore, vendors see a need to provide drag-and-drop functionality in order to compensate the weakness of scheduling logic.
But, the drag-and-drop functionality can also create a problem for the user if the scheduling software cannot properly assess and address the ripple effect and the resulting constraint violations. The software must at least reveal the effect on the schedule and give users an option to undo the drag-and-drop operation. A single drag-and-drop operation may sometimes necessitate many such operations for eliminating the negative impact on the schedule.
The goal of any developer of scheduling software must be to minimize the necessity to perform drag-and-drop operations on Gantt chart. It can be achieved with the help of versatile and flexible scheduling paradigm and rigorous scheduling algorithms. Scheduling software based on advanced finite capacity scheduling logic generate high quality schedules without causing overloading of any resource (when overloading is not tolerated). Such tools do not create a need to perform a lot of drag-and-drop operations on Gantt chart to repair a schedule or improve schedule quality. One of such powerful tools is our software, Schedlyzer.
A frequent need in a scheduling tool to perform drag-and-drop operations on Gantt chart is a strong sign of the weakness of core scheduling logic in the tool.
The author, Dr. Prasad Velaga of Optisol, LLC has 20 years of experience in developing powerful, scientific scheduling solutions which are regularly used by many complex, order-driven, high-variety production systems.
Co-Owner at Simply Effective Solutions
5 年Drag and drop demos so well though :) Completely agree that in practice it is a curse (unfortunately users don't realise that until after they have made the purchase. I have often thought about it ... Why does a feature demo so well and is often the reason people make a purchase and yet it turns out to be a curse in practice).? At the end of the day users want "flexibility" and even though algorithms can do everything for them it is too much of a black box. They really have no idea how the program sequenced their jobs even if it is optimum. If the planner can't defend the schedule from questions coming from the production managers it will be worthless as the production managers will do what ever they want on the shop floor. (If you can't justify a decision it will give them license to do what they think is best. The optimum schedule is now worthless). Their is a good compromise though by allowing users to sequence bottleneck resources first and then automatically scheduling around that (i.e "layered resource" scheduling). You can give them a first cut of the sequence but we have found it is critical to allow users to override the sequence to meet their specific needs for that day before automatically scheduling everything else.