Buffers are a novel concept in LBMS. They function similarly to CPM lags during planning but they can be absorbed and are similar to float in controlling and forecasting phase. Buffers give superintendents time to react if there is a problem with predecessor’s production. They also make the project easier to control because there are fewer operations ongoing on site, they are of long duration and continuous and start dates are spaced out. Subcontractors work faster if there is work ahead of them so any time lost will be regained during production. That is the theory and there is evidence that it works (both empirical and simulation, for example try my production control game) but why are buffers not being used in LBMS schedules?
The main reason seems to be unfamiliarity with the concept. On the surface it seems that buffers delay the project and that is indeed the case during planning phase. However, in LBMS schedule optimization crews are added to balance flows and sequences are optimized. This typically results in a 10-20% compression in project schedule compared to the CPM start point. Durations in each location are typically shorter. This is fine because the subcontractors can work faster if they have the space and all constraints have been removed. Buffers are required to make sure that the subcontractors have the location available when they get there. Instead of delaying the project, buffers take it back closer to the original duration. For example, if LBMS is used to compress 20% of project duration based on continuous flow and subcontractor productivity data, buffers can then be added back to end with just a 10% shorter schedule.
Caption: One best practice is to place buffers between trades which cannot overlap; this example illustrates adding buffers between drywall, pipework, ceilings, and finishes.
Where should buffers be planned? Between any two trades which would interfere with each other when working in the same location or when the predecessor must be completed before the successor can start. Typically it makes sense to buffer tasks which are continuous, have high man-hours or require large lay-down areas. Important inspections should always be buffered. The buffer should be added AFTER the inspection to prevent delay caused by failed inspection. Buffers are not required for tasks which do not take a lot of space and can easily happen concurrently with other work (e.g. door installation).
Caption: Another best practice is to apply a buffer after important inspections. This way the buffer protects the schedule against a delay for the rework.
To make buffers easier to use, it makes sense to create a report of all buffered tasks. A new task should be started when the predecessor has been completed in the location and the buffer time has elapsed. This may lead to slight delays in the first location but if the work continuity is preserved and the task is of long duration, it is easy to catch up during production by just increasing the crew size by a few people.
To learn more about Location-Based Management System and how it is deployed in Vico Office Schedule Planner and Production Controller, please see:
Video Training Series: Schedule Planner
Video Training Series: Production Controller
And be sure to catch up with the other best practice articles in this series:
Part 1: Beyond Start Dates
Part 2: Get the Subs Involved
Part 3: Manpower and Suspensions
Part 4: Control Actions
Part 5: Planning the LBS
Part 6: Clarifying Scope
Part 7: Look-Ahead Plans
Part 8: Running Parallel Schedules
Part 9: Resource Graphs
Part 10: Production Rates and Location Sequence
Part 11: Using 5D with LBMS in Subcontractor Meetings
Part 12: Deliveries and Lay-Down Areas
Part 13: Starting As Early As Possible Will Hurt Your Project