Sunday, October 3, 2010

Research Results

As I explained in a previous post, we get funding from somewhere and conduct research. Like any other kind of work, research too produces results. In this post, I will explain the main types of research results.

First, some of the more obvious results for different fields of research:

1. Computer science: large sales of coffee and computers.

2. Bio-science: Piles of dirty lab-coats.

3. Humanities: Books that can be used for weightlifting.

Now, the less obvious but more important results.

1. Products:
Someone said "research is the design of products that are used in the future." But only a few researches directly end up in products. That is because research outcomes are not always good enough to be used as products. Let's take an example.

Suppose researchers developed a device to automatically destroy terrorists. The device will have a camera that will take photos of people. Then a computer will analyze the photos in a few milliseconds and determine if the people are terrorists or not. If it determines that a person is a terrorist, it will shoot him/her with the gun attached to it.

The researchers will run a long experiment and show that it will detect 80% of terrorists that are known and let the other 20% go (but then, this is probably 8 times better than allied forces in Iraq and Afghanistan). It will also kill about 25% of the innocent people that its camera sees (that is already 4 times better than Israeli forces in Palestine, I assume). Finally, terrorists will soon find that a flashlight will completely crash the thing. So much for products :-p


2. Patents:
Researchers do create lots of patents. Some workplaces measure their performance by the number of patents filed per year. Some patents generate a lot of money some time later. But most of the patents contain things like "I found that if a middle aged fat guy stands on a manhole cover that is in front of a gasoline stand, he will get a mild headache." Very useful, isn't it? But then, patents also help us a lot by filling pages of our resumes and getting us through the yearly performance appraisals.

3. Startup companies:
Sometimes, a research result looks promising but big companies do not want to use it. Some of the more brave researchers try starting companies using such results. Google is one successful example. The failures just disappear in less than two years. But we admire those who fail for trying to do some real work.

4. Cool exhibits:
I guess these are more common to Japan than other countries. Some research companies develop things like cute robots to show the latest results of their research. "Aibo"from Sony, "Asimo" from Honda, and the trumpet-playing robot from Toyota are examples.

5. Publications
When researchers get reasonable results from their work, they write "research papers" about those results using long passive-voice sentences. They then submit the papers to conferences or journals. After evaluation by who are called "reviewers", a limited amount of papers are published. People who are interested in long passive voice sentences can read them, if they want to.
I admit that it sounds much less entertaining and useful than the previous types. But then, number of publications is the main performance metric for researchers. So, some of them do crazy things to make this number higher. I will write about them later, for your entertainment :-p.

No comments:

Post a Comment