Thursday, August 04, 2005

A Real World Efficiency Problem

So, here's a question that I'd like input on. It has to do with efficiency, measured in terms of total time for project completion. Let me describe the problem and some pros and cons of a couple of paths.

We have 25 participants who need to go through a data collection process. The data is being collected under 6 different environmental settings. The environments are being created for us, since the data collection is taking place inside a large temperature / humidity chamber. In this chamber, there are 3 data collection stations: A, B, and C. Data collection on Stations A and B takes approximately 4 minutes of time on each. Data collection on Station C takes approximately 2 minutes of time. Some of the environments are harsh, meaning that prolonged exposure could cause some serious discomfort. Assume that a person feeling serious discomfort cannot work as quickly to complete data collection. Each participant needs to have their data collected on all three stations at all 6 environmental settings. When this occurs, the data collection project is deemed complete.

Some people will take longer than others, sometimes twice as long, on certain stations. However, we do not know how fast each person will be a priori. Only when all the data collection has been completed at a given environmental setting will the chamber be re-configured to simulate a new environment. Adjusting the environment takes approximately 15 minutes of time.

Now that I've furnished you with the background information... I'll propose two different ways to handle this data collection process, as well as provide pros and cons of both. I'd like to hear what you think, and what other adjustments could be made with the end goal being the total time it took to complete all the necessary data collection.

Path 1: Line up stations in order as Station C, B, then A. Form a line outside of the chamber. The Person #1 goes to Station C. When he completes it, he moves onto Station B, and Person #2 goes to Station C. As soon as each person has finished at a station, he will move to the next spot as soon as it becomes available. So, we basically form a queue and move everyone serially through the three stations. We repeat this for each environmental setting.

Pros: Very organized method -- should be a single pass and we're done. Also, less overhead due to participants not knowing exactly which station they should go to, or if they've already finished on one.
Cons: Less efficient use of each station if there is a slow person causing a bottleneck. Could increase the average time of exposure to each environment, causing additional slowdown.

Path 2: A line will form outside of the chamber, just like in Path 1. However, Persons #1, #2, and #3 will enter the chamber. The first one that completes the data collection at his station will wait until a new station opens itself up. He will immediately fill any open station with which he has not yet collected data. Only after a person has collected data on all three stations in the chamber will he leave the chamber to allow another person inside. At all times three people at most will be inside the chamber.

Pros: Seems to keep station utilization at near 100%.
Cons: An organizational nightmare. Could force us to re-collect data when participants miss a station or collect data on the same station twice. Bottlenecks will still cause similar problems as Path 1, but to a lesser extent.

Which of the above paths do you think would be the better one? In Path 1, does the ordering of Stations A, B, and C matter? Practically speaking, is there that much time to be saved by taking Path 2... data re-collection can be quite time consuming. What if we ignored potential data re-collection time... how much time are we likely to save by forgoing organization? What other factors should be considered?


Anonymous said...

Definitely Path 2. It's worthwhile to undergo the trivial organization/logistical planning (e.g. supplying each person with a form with 18 checkboxes, so they can check off the stations they've visited, for the 6 different environmental conditions) and save time for everyone involved.

Anonymous said...

I also think Path 2 is better. However, I do think that when one person is finished and there is a 15min wait time as the environments change, that person should move on to line up at the next station. As Jim pointed out, giving them a checklist will be helpful so they don't have to rotate in an organized fashion, they should just line up to where they can enter quickest.

There should be one part of a group that is just standing to the side, and one group rotating through. Upon completion, the guy will yell over to the next guy to come join the group entering the stations to take his place. That would save some confusion about who goes where b/c you will never have all 25 people lining up and bumping into everyone.