
Welcome back to our series on Business Process Automation. In this post, we focus on enhancing business processes, particularly in the context of our hairdressing salon case study, using simulation and process mining tools.
Recap: Identifying the Bottleneck
In the previous installment, we used WoPeD to simulate the business process of a hairdressing salon. The results revealed a significant bottleneck at the payment stage, where customers experienced lengthy waiting times. This issue demanded a closer investigation to determine the root cause and implement effective solutions.
Leveraging Process Mining for Deeper Insights
To further analyze the bottleneck, we utilized WoPeD’s capability to export simulation logs in the XES format, a standard for process mining. These logs were then imported into the process mining tool ProM, which allowed us to visualize the salon’s workflow as an animated process model.
Other process mining tools, such as Disco or Celonis, could also handle XES files.
Analyzing the Animation in ProM
The process model derived from the logs provided crucial insights. Each yellow bead in the animation represented a customer. When beads moved slowly before a task, it indicated either ongoing service or waiting time. For example, the task “cut hair” included a 20-minute service time, whereas the task “pay” involved almost entirely waiting time, with only a two-minute service duration.
Implementing Changes: Assigning Bruce to the Till
To address the bottleneck at the payment stage, we modified the process by assigning Bruce, the salon owner, to handle the till. In WoPeD, we adjusted the role of the “pay” task from “hairdresser” to “owner.” Bruce was already included as an “owner” in the WoPeD model, simplifying the transition.
After running a new quantitative simulation, the results were encouraging. The average waiting time before the “pay” task was nearly eliminated, with Bruce now spending about 30% of his time handling payments.
Verifying Results with ProM
To confirm the improvements, we revisited the animation in ProM. The updated process model showed that beads moving to the “pay” task now exited quickly, while those continuing to “color hair” experienced a more typical flow. This verified that the bottleneck had been successfully resolved.
Conclusion
By combining business process simulation and process mining, we effectively addressed a critical bottleneck in our hairdressing salon’s workflow. Assigning the payment task to the owner eliminated excessive waiting times, improving overall efficiency and customer satisfaction.