Inspire

Adorned on a landscape background, sitting in posters across classrooms, a Thomas Edison quote reads "I have not failed. I've just found 10,000 ways that won't work."¹ It is an inspiring thought that is very easy to say in hindsight. But what if you are on trial #1,912 or #3,504 and all you have faced thus far is failure?

Failure is an absolute certainty, especially when trying to learn something new or tackle something innovative. Often you must fail over and over again until maybe succeeding. If you knew that your failures were bounded at 10,000, then that would actually simplify the problem. But the number of failed attempts we will experience is rarely known and is usually unbounded. There is no limit.

But that’s OK.  Fred Brooks (a huge, huge hero of mine) is quoted as saying:

“You can learn more from failure than success. In failure you’re forced to find out what part did not work. But in success you can believe everything you did was great, when in fact some parts may not have worked at all. Failure forces you to face reality.”²

But even if you expect to fail and fail a lot, and understand Brooks’ wisdom that failure is natural, important, and part of the learning process – it does not soften the blow. It can knock you off your feet. It can make an “adult” college student cry on her apartment couch so long she gets a headache and her roommate cajoles her unto eating a lot of milk and cookies.³

So we have established Theorems 1 (Failure happens) and 2 (It hurts). The proofs are derived from life experience and therefore are left as an unfortunate exercise for the reader.

But there is something we can do to mitigate the consequences of repeated failure.  I'm a research roboticist (in addition to being a student) and one of my favorite non-technical quotes is the following:

'Benjamin Franklin is often cited as saying, “If you fail to plan, you plan to fail.” In robotics, if you do not plan to fail, you are failing to plan. Uncertainty and unexpected events are part of robotics and designing resilience into a system is how to address this reality.'⁴

This provides us with the corollary to our previous theorems: plan for failure. We need to design resilience into our learning process and into ourselves. If you have six weeks to build a robot⁵, build slack into schedule for when things go wrong. If a project seems to blow up in your face, use milk and cookies to cure your headache and get a good night’s sleep.  Then, the next morning, write a post-mortem about what worked and what didn't.

So in learning to succeed, we fail, and that can hurt. But in order to succeed, we must learn how to fail. How to fail such that we get back up. How to push pass trial 999 to get to trial 10,000. We have to learn how to succeed at failing.
 

1. Bypassing that this quote is disputed, most of us have probably seen it in this form.
2. Kelly, Kevin. “Master Planner: Fred Brooks Shows How to Design Anything.” Wired magazine.  Web. July 28, 2010.
3. The most recent incident was due to a missed conference paper deadline. Human-in-the-loop machine learning proves to be incredibly tricky. But we haven't given up - I'm still working on it!
4. Johnson, Matthew, et al. "Coactive design: Designing support for interdependence in joint activity." Journal of Human-Robot Interaction, 3 (1), 2014 (2014).
5. This is clearly hypothetical as no one would actually do something that outrageous.
 

Rachel Holladay is an undergraduate student and researcher at Carnegie Mellon University and is a FIRST Alumna, Mentor, and Dean’s List award winner. Read Rachel's bio.

If you have an inspiring story or piece of wisdom that you’ve picked up through your experiences in the FIRST community, please reach out to us at inspire@firstinspires.org and inquire about becoming a guest contributor for Inspire.