Recently a student asked, "What advice would you give to yourself back when you where just finishing college?" It's a great question, and after a fun discussion we've come back to something very similar to my first, gut instinct response:
Do something you are passionate about. If you are passionate, then you will work hard and learn. If you work hard and learn, then every form of success will follow.
Get to know yourself and your blind spots. No one is good at everything. Get to know what you are not good at, and learn ways to work with people who can compliment your weaknesses.
Make mistakes and learn from them. Make as many mistakes as you can. Find a way to double your rate of failure, because as long as you are learning you are getting closer to your goal that much faster.
Take time to teach. Even if you don't become a teacher, become someone who listens, understands and teaches. Teaching is hard because the better you know the material, the harder it is to relate to the student.
Cognitive Load is a concept in psychology where a person has a more difficult time performing a given task if there is too much going on. For example, it is easier for me to learn about history if you present the information to me in English and don't ask that I do math problems at the same time.
In test and measurement, the same phenomena exists. If you need to perform a certain test, and running around the lab reading and adjusting all of the equipment takes all of your time, energy and attention, it is unlikely that you will notice an important discovery which may be hidden in plain sight.
Software Architectures: State Machine vs Message Handler
Wednesday, 16 November 2011 17:22
This week a student asked about the fundamental differences between a state machine and a message handler. This is a particularly good question as these two software architectures are very common in instrumentation, controls, test and measurement applications.
First off, congratulations to this student for asking the question. If you are struggling to select which architecture is best, then it's very likely that they will both work well. More importantly, simply by using a software architecture this student will ensure that his code will be far more organized and far easier to debug, maintain and expand than code without an architecture. An additional benefit is that it is fairly easy to convert between state machine and message handling architectures, so if the student selects one and future development shows it wasn't the best choice, the change to the "correct" architecture is usually straight-forward.
It is important to have the software architecture resemble the system. This allows the most flexibility going forward and makes everything work out best. It is up to the student to understand his system, but to make the best choice we also need to know about the architectures:
NASA announced on 11/29/2011 that Angstrom Designs has won it's first NASA Small Business Innovation Research (SBIR) proposal. The Angstrom Designs NASA SBIR award win is due, in large part, to a solid partnership with established business in anticipation of commercialization. The SBIR team at Angstrom Designs is proud to serve NASA and our country.
Stay Up To Date: Subscribe to the Bi-Weekly Newsletter
Thursday, 08 December 2011 22:40
Get Blogs, software updates, new applications and more in your e-mail inbox. Op-in to our newsletter. It's spam safe and you can unsubscribe at any time.
LArVa in Test & Measurement World
Wednesday, 09 November 2011 00:04
Here is a short article about Arduino and LArVa, as written up in Test and Measurement World:
A former student, Christian, asked one of our senior systems engineers, Casey, about generating lots of intellectual property for a start-up. The discussion is below, but it comes with the disclaimer that very, very rarely should patents even be considered before a large amount of market validation has been done.
When you were writing / publishing a patent every 6 weeks, what was your schedule? What helped you move more quickly? Did you work with any (lawyers / patent writers) that made the process easier? I mostly see a brute force way to get things done. Did you discover a method that helped to generate material faster?
The board wanted an aggressive patent schedule, so I proposed a 10 week spin. Each patent took longer than that, but we put one in the front of the funnel and pulled one out the back every 10 weeks or so. To accomplish that I hired some help to organize and drive the process. We then broke the process into 4 stages: ideation, sorting, distillation of the idea and final authoring.