Operations Excellence: The Critical Path

In a recent experience working with a client, we were process mapping the timeline for bringing a product to market. As we were working to develop the Gantt chart of timelines for the key phases of this process, the client offered several qualifications. "Depends" was a common answer to our questions about how long a process step takes, or when it might start relative to other steps of the process.

Marc Puich

Managing and improving complex processes with variability in the work that is executed by them is challenging. A typical first step when analyzing a complex process, is to develop a Gantt chart to identify the critical path. This helps identify steps that are the priorities from a scheduling and process improvement perspective. All the other shorter steps of the process that happen in parallel with these critical-path processes have more flexibility in their completion time, assuming they are completed before the end of the critical path items.

But what if the critical path shifts depending on product or other constraints? Now you have a situation where the participants in this complex process may not be aware that they are now on the critical path. The typical outcome of this situation is missed due dates and poor operational performance.

An awareness of the critical path can sometimes be a risk. Human nature and large workloads result in individuals waiting until the last moment to complete their tasks within their lead-time targets. However, as we know, changes happen (rework, critical information is unavailable, etc.), and now this non-critical path item has delayed the entire process.

An example of this is the product release process. Let's assume that it is supposed to take 50 days to release a batch of product. The Gantt chart of this process (Figure 1) shows that laboratory testing, which is assigned 48 days, is the critical path process. Investigations have historically averaged 24 days to close, and so an investigator waits until day 24 from the end of production to begin work on this particular deviation. What this investigator forgets is that 24 days is an average and this product is new, and investigations for new products typically average 35 days. The original assumptions, which governed the investigator's work, no longer hold true and the lot will take longer than 50 days to release.

Figure 1. An example of a Gantt chart for batch release

COMBATING COMPLEX PROCESSES

How can companies combat these situations? Two main approaches can help eliminate this problem: project management and using start dates instead of end dates.

Project management

In environments like the one described above, with variability in the products and standard times, someone needs to review the overall process and ensure that all contributors are aware of their due dates. Also, for longer processes, there should be intermediate milestones so the process owner can ensure that progress is being made. Organizations rarely put a project management structure into batch release. It isn't until the process is near completion that the delays are identified.

Start dates, not end dates

Another strategy is to focus on the start time of process steps, not end time. In order to prevent delays, goals should be set to start activities within X days of when the precedent activities are completed. A completion target time should also be measured. This way, simple or non-critical path processes are started as soon as is practical, with ample cushion to end without impacting schedule. A follow-on effect of completing activities early is to identify and address unforeseen issues that could contribute to delays in the future. All contributors to the process need to have a clear understanding of their capacity and see these work activities as a priority, regardless of where they are on the critical path.

In the challenges that biopharmaceutical companies regularly face in areas such as technology transfer, product release, change control, and validation, identifying the appropriate structure, management, and targets and prioritizing them well can make a big difference in ensuring critical lead-time goals are met.

Marc Puich is a partner at Tefen, Ltd., Foster City, CA, 650.357.1120 x117, marc@tefen.com