NASA's Discovery: Too Much Information?By Tom Steinert-Threlkeld | Posted 2005-07-27 Email Print
How Real-World Numbers Make the Case for SSDs in the Data Center
NASA pointed 110 cameras at the shuttle launch and still doesn't know what fell off. IT execs face a similar problem every daytoo much information, too few answers. New studies show how to filter data for answers, and how to focus your own messages
110 cameras, on the ground, on the fuel tank and elsewhere on the shuttle. Two trailing weather planes, taking photos through a telescope. 90 sensors on the wings, taking note of the impact of any objects. Radar, to track movements.
That's the quick summary of technology NASA is using to try and prevent any disaster on its Discovery shuttle. So is it surprising that it found some stuff falling off the sides of the space vehicle? Or that it will take days to sort out what the seriousness of the events were? When can you overload your organization with too much information? >
NASA's belief appears to be: You can never have too much information. But studies indicate that the human brain can only sort through and compare a few variables at a time.
University of Queensland psychologists, for instance, pushed 30 academics and found their mental limits were hit when four variables were compared at one time. That theoretically becomes a bottleneck in evaluating complex problems.
The problem facing NASA really is how to deal with all the information from its camera, sensor and radar systems. So far, it's easier to capture data and even knowledge. But it's a lot harder to build into software the processes that engineers and scientists use to evaluate problems.
You might not have to worry about vehicles blowing up with your employees inside (although the top brass at the MBNA credit card outfit recently all went down in a helicopter in the Hudson River). But if you have a complex business you're trying to monitor and lots of data coming in that you'd like to inform you about what's going on, here are basic steps you can use to organize how you analyze it. Because the variables are so great, organization to organization, department to department, situation to situation.
- 1. Identify four criteria or questions that you want the information to answer.
- 2. Establish the sources of the data you need (or create them).
- 3. Sequence the steps you will use for analyzing information.
- 4. Choose a technique for analyzing the information, whether it is costs versus benefits, effects on a critical path, a decision tree or other methods of bringing order to complexity.
- 5. Benchmark the results, so you can see departures from the norm.
But the most important principle might be: When you get the analysis done, present it simply. After the Columbia shuttle blew up, the analysis of wing damage was buried and almost lost in a jam-packed PowerPoint presentation. Here's an analysis of the key slide.
Engineers involved had obscured understanding by committing information overload.
Analysis: To make a point, make just one point