Re-Thinking Critical Chain Scheduling
|Robert Newbold - CEO of ProChain Solutions|
|Three years ago, at the TOCICO conference in Bad Nauheim, Rob Newbold argued that feeding buffers should be eliminated. Now he will present the full set of changes needed for critical chain scheduling to become more transparent, intuitive, and powerful; thereby making critical chain adoption easier and more effective. In the process, he will challenge a few sacred notions.|
At a high level, the power of this new technology is to make critical chain scheduling easier to understand and teach, more transparent and intuitive, and hence more likely to be adopted widely. The main changes are as follows:
1. This technology addresses the myriad contradictions and workarounds embodied in the canonical or "traditional” critical chain. These contradictions and workarounds limit people’s willingness to learn and adopt critical chain. For example:
- a) Eliminate feeding buffers, using alternate mechanisms to determine task gating and to account for merge bias ("integration risk”). This is per my Hyde Park presentation in Bad Nauheim.
- b) Make the critical chain equivalent to all tasks responsible for the duration of the project.
- c) Update critical chains, gating times, and integration risk calculations whenever the schedule is updated, not just on a full reschedule.
- d) Incorporate the highest variability tasks in calculating the project buffer size, even if they’re not on the critical chain.
- e) Add tools to analyze how and why the buffered project duration differs from the critical chain duration.
2. People try to focus on the critical chain, but two problems are encountered:
- a) People have difficulty understanding and analyzing feeding buffers, task gating, and integration risk.
- b) People waste time trying to manage feeding buffers even though the benefit is questionable, especially since feeding buffers often become irrelevant as a project evolves.
- c) Multiple endpoints, for example for multiple deliverables, must be dealt with through special mechanisms such as contractual milestone buffers.
3. There are many policies, norms, measurements, behaviors, and processes used to bypass the limitation. I mention here a few that correspond to the items in under answer 2 above.
- 1. When several chains are mathematically critical, "just picking one” in order to promote focus actually has the opposite effect. Non-critical chains are more likely to become more critical, and lack of confidence in the schedule means that people are less likely to focus on the few critical tasks.
- 2. When there are significant changes to the project, the critical chain quickly becomes obsolete.
4. Once the new technology is in place, various new policies, norms, and behaviors can be implemented, such as:
- a) People are discouraged from asking "why” questions about schedules, because those questions can be very difficult to answer. For example: why does this buffer have this duration? Why is that task scheduled to start tomorrow? Alternatively, when important questions must be answered, those answers can be difficult to obtain and to understand.
- b) People ignore feeding buffers. As a result, we lose track of the very real safety time that’s needed to protect against merge bias or "integration risk.”
- c) We create extra types of buffers, or excessive pressure to manage only one deliverable.
- d) People resort to illogical workarounds during execution to try and focus on the most critical tasks. Those workarounds include excessive efforts to try and keep the critical chain from changing; ignoring the critical chain, and instead using other indicators after the schedule is activated; and/or rescheduling.
5. The new technology has been designed with the new rules in mind, so the main changes are in how we bring people up to speed on the technology. The new rules mean that it is simpler to teach and use the critical chain methodology. That means the time and cost of the training and learning is less, which means the barrier to entry – for individuals and for companies – is lower.
- a) Encouraging all questions, because they can quickly be answered. While often the answers aren’t directly relevant to project performance, they build a sense of confidence in project team members in the technology and in their understanding of it.
- b) Truly comprehensive schedule analysis. Schedule analysis can be grouped into three areas, of which the first is by far the most important:
- Analysis of the "critical chain,” which includes all parallel chains. Does it make sense? What are the risks? How can we speed it up?
- Analysis of buffers and why they are what they are, what I call the "variability chain.”
- Analysis of time inserted to protect the critical chain (formerly done through feeding buffers).
- c) Less frequent and less disruptive rescheduling, because feeding buffers and critical chain no longer become obsolete.
- d) The decision to reschedule is much simpler: do we need to revisit the project completion date? Changes to feeding buffers and the critical chain no longer require rescheduling, and instead are kept up to date via the normal schedule update process.
6. We have found no difficulty causing the change for clients. While continuing support of the traditional scheduling approach gives people comfort, once people try the new approach they don’t change back. Causing the change more broadly in the TOC community would require a few pieces.
- a) Software vendors need to understand the benefits of incorporating the new concepts. Fundamentally, those benefits come from (a) decreasing the barriers to entry and (b) increasing the ratio of software dollars to consulting dollars.
- b) Practitioners need to understand that the current approach is simpler and better. They can spend less time on obscure technology and more time helping people get more throughput. We have found that some practitioners are wedded to the "traditional” approach and may never change. However, so far the majority have had no problem with it.
Robert Newbold is founder and CEO of ProChain Solutions and one of the world’s leading experts on implementing Critical Chain project scheduling and management. He has over thirty years of experience developing process improvements in various fields. Rob is a frequent writer and speaker and holds degrees from Stanford University, SUNY Stony Brook, and Yale University. He is the author of the books The Project Manifesto (ProChain Press), The Billion Dollar Solution (ProChain Press), and Project Management in the Fast Lane (St. Lucie Press) and is a contributing author to Theory of Constraints Handbook from McGraw-Hill.