Computer Simulation as a Tool for Reorganization and Automation Planning

In the last issue of JALA 3, No 2, pp 31–33, 1998, we published a paper on computer simulation. The figures and legends of two different versions, submitted to the journal appeared incorrectly. We therefore publish the figures with the correct legends and a short summary once again.

The basis of our experiments was a simulation engine under DOS, called Simlab® (1), which we have recently developed further to make it useful as a consulting tool. Version 3.0 of the program includes Windows modules for easy data capturing and on-the-stage presentation of the results (2).

We offer simulation services to clinical laboratories and industry. A typical study takes about ten days. For more information you may visit our website at http://www.trillium.de.

The figures on this page show selected results of two pilot studies, conducted in the US and in Germany. Each single experiment (“scenario”) created several 100 data for staff and equipment utilization, turnaround times and bottle-necks in the process.

In 1996, the University of Virginia made plans to install a sample preparation front-end from Beckman-Coulter. Our simulation study predicted that the frontend would:

• Be reasonably utilized under the current workload conditions (Figure 1)

• Decrease the average utilization of the seven technicians in the accessioning area by 5% to 55%, depending on the hour of the day (Figure 2)

Figure 1:

Predicted utilization of a front-end automation system at the University of Virginia

Figure 2:

Predicted utilization of seven receptionists of the University of Virginia before (blue line) and after (pink line) installation of a front-end automation system.

In 1997, the German Heart Center in Munich planned expansion of its laboratory due to an expected doubling of the workload. The simulation study said that:

• Doubling the workload would create major delays (Figure 3) and bottle-necks (Figure 4)

• The original performance of the laboratory can be achieved by combinations of hiring more people and/or buying new equipment (Figure 3)

Figure 3:

Balancing turnaround times (left) against fixed costs per test (right) at the German Heart Center

Figure 4:

Queue length analysis at the German Heart Center in Munich after doubling the workload (yellow) and after reorganization (blue). The figure shows numbers of tubes queuing up in front of manual (yellow) and automatic (blue) aliquotting.

The most cost-effective solution was hiring one person and buying an aliquoter and a consolidated coagulation instrument. This restored the TAT from 153 to 58 min.