Resourcing Schedules – A Conundrum

I have recently been forced to think about the value of incorporating resources into schedules. At one level it’s not too hard to do, but is it useful?

From one aspect, it is impossible to schedule at any level without the active consideration of resources. Resources do the work in a given time and changing either the quality or quantity of the resource has some inevitable impact on duration. Consequently, it is critical to know the resource assumptions used in planning to validate the schedule and more importantly understand deviations from the plan during the execution of the work.

Generally what I mean by term ‘considered’ is the basic need to know the resources needed to undertake the work on every activity:

• At the feasibility stage big picture tied to the strategy for the project.
• At the contract stage to determine which tasks are the responsibilities of what contractor/subcontractor.
• At the weekly level, the supervisors need to know who is working where and when.

These decisions also need to be recorded and monitored. How much detail is recorded in the scheduling tool and what scheduling functions are used though is an altogether different question – this I refer to as ‘quantitative’ resource analysis.

Consideration is not the same as quantitative analysis within a scheduling tool. Quantitative resource analysis requires answers, or assumptions to be made, about a range of uncertain issues. Some of the nearly insoluble questions include:

1. There is no direct ‘straight line’ correlation between resource quantities and either task or project durations – there is a complex ‘J’ curve relationship and in some circumstances a negative correlation. For more on this see: The Cost of Time or for a more learned approach, The Mythical Man Month by Frederick P. Brooks Jr. originally published in 1975.
2. It is nearly impossible to define skill levels for people who will be employed on a project at some time in the future but we know a skilled worker can be far more productive than an unskilled worker. The skill of the worker changes the production rates and consequently the durations.
3. The other issue is the degree of motivation/moral of the people – a highly motivated team will always accomplish more than a ‘business as usual’ team and both more than a de-motivated workforce. Therefore the question of management and more importantly leadership also influence resource performance and therefore durations.

These unanswerable questions are complicated by the fact all scheduling software fails to optimally level resources . Basically the tools get it wrong the only question is how wrong: some are not too bad others unmitigated disasters. Resource scheduling needs both knowledge and common sense – no software applies common sense yet. But we have to plan resources – they need working space, accommodation, etc. And resources are the source of all cost expended on the project!

Another really interesting factor is the emerging understanding of the interaction between the schedule and the behaviour of people. IF the people believe the schedule represents a realistic approach to their work, they will (and do) modify their behaviours to conform to the schedule to be seen as successful. Obviously if resources are included in the schedule it is far more credible than if they are not. This was touched on in Scheduling in the Age of Complexity (read from p19 – the rest is not relevant and it’s a horribly long paper…. with a bit of luck this may turn into a book in a couple of years….).

So in conclusion I would suggest, consideration of resources is critical, as is having some form of method statement; together they dictate the planned durations of the work.

However, whilst using scheduling tools to calculate and level resource demands is useful, and can help gain valuable insights, you need real skill on the part of the scheduler and the right tools to achieve sensible results.

My feeling is the value of the process to the development of a realistic and achievable schedule depends on the circumstances of the project. Probably the biggest determinant of the value of quantitative resource analysis is the ease of adding to or reducing the resource pool. If this is easy, rudimentary quantitative analysis is all that’s needed, if any. If it is difficult to quickly change the resource pool far more rigour is required (eg, developing remote area mine sites in Australia). The quantitative analysis will still be ‘wrong’ but it is important to reduce the level error as much as possible.

This is a complex issue – what are your thoughts?