Improving your lab with simulation

From LabAutopedia

Jump to: navigation, search
Invited-icon.jpgA LabAutopedia invited article

Authored by: Dr. Michael Allen, Lean Sigma Lab


Simulation is the technique of building a model of a real or a proposed system, such as the day-to-day operation of a laboratory, a manufacturing facility, a new robotic platform or even the behaviour of a portfolio of research activities over an extended period. Almost all systems can be modelled, with the model taking into account resources available, workload, and the normal planned and random fluctuations of work and resource availability.


The impact of simulation on your lab or R&D process

Simulation is a very powerful technique as it allows experimentation in a virtual environment. Experimenting with live processes may be costly, risky or stressful to those involved. Simulation can be utilized in a number of ways to improve the operation of a laboratory. Following are a few examples.

  • A simulation model may be investigated by overloading it with excess work to expose bottlenecks, or running it on reduced resources to better understand impact of these conditions on output and cycle time.
  • Careful simulation model analysis may suggest process improvements and the modification of work plans, without the need to interrupt the current facility. When changes are implemented they are done so with confidence and a minimum amount of risk.
  • Simulation models will help to engage stakeholders. Visualizing a model of a new facility or proposed changes to an existing process will generate confidence that proposed modifications will result in a successful outcome. Stakeholder support will be more easily earned and maintained throughout the improvement project.
  • Simulation can be a key step in resource planning. With a simulation model it may be possible to accurately gauge capacity, throughput and cycle times for complex systems, leading to the deployment of a measured FTE/equipment resources that match desired outcomes.
  • Simulation models can be used to demonstrate how a system will cope with unexpected demand. For example, models can be used to identify ways in which redundancy in other systems are released when the process is stressed beyond typical operating capacity.

A simulation example

Let's look at a very simple example. We have an analytical laboratory that receives samples and tests each sample in two assays. Samples arrive randomly throughout the day. The samples are split (which takes about 4 minutes), prepared for the tests (about 6 minutes per test), tested automatically (about 7 minutes per test) before the results for each patient are collated and e-mailed out (about 3 minutes). The process (as the tests may be carried out in parallel) requires about 22 minutes for one sample to pass completely through. When we draw the process as a simulation diagram, it looks like Figure 1.

Figure 1.

Unlike a Visio or PowerPoint diagram, the process diagram in Figure 1 also runs a simulation. Work is put through the simulated process steps in accelerated time. Sample queues and resource utilisation are monitored throughout the simulation. Upon completion of the run we can analyze results (Figure 2). We see that queues grow and shorten throughout the simulation and that the average cycle time (including queuing) was 5 hours (306 minutes). Although the slowest step in the process is identified as the automated test we can see that lab staff utilisation was 97%. Consequently, the factor limiting performance of this system is not actually the slowest step (the automated test), but the lab staff. We look to either the process step or resource that has the highest utilisation to find what is constraining the system.

Figure 2.

Rather than investing in more equipment, which appears to be the slowest step and could easyly be mistaken as our problem, we look to reduce time spent by the staff. We could, of course, invest in more staff, but our first step should be to try to improve the process. Running the simulation again we find that if we reduce the set up time to 5 minutes (a reduction of only one minute) we reduce cycle time from 5 hours to 2 hours. By contrast, doubling up equipment in the model would have only reduced cycle time by 30 minutes to 4.5 hours!

Simulation can help you understand and analyse your current or planned system. When you understand your system well you can quickly focus on the areas that will truly enhance process performance!

External Links

Introduction to Simulation

Click [+] for other articles on  The Market Place for Lab Automation & Screening  The Market Place
Click [+] for other articles on 
Project management(1 C, 9 P)
The Market Place for Lab Automation & Screening  The Market Place
Click [+] for other articles on 
Quality(6 P)
The Market Place for Lab Automation & Screening  The Market Place
Click [+] for other articles on 
Process management(3 P)
The Market Place for Lab Automation & Screening  The Market Place