What Is Innovation?

Before heading out to the ASME DSCC conference in Washington DC last month, I was trying to answer this question.  When you’ve visited over a hundred academic engineering research labs across the US and Canada, you start to recognize some patterns - where some researchers struggle and others thrive.  

The popular understanding of innovation is something that is borne from a eureka moment - maybe someone sitting on a stone and thinking up the next Facebook or hitting their head and coming up with a Flux Capacitor.  

However, for many of our clients involved in engineering research, this is hardly the case.  It takes many years of toil, guiding graduate students, dealing with lots of challenges, and only the faintest prospect that what they’re doing is going to pay off in a big advancement.  

If we define innovation as expanding the physical or thinking capability of humankind, most of the work involved in research does not cause innovation.  It’s only at the very end of the work that some small expansion of human capability happens.  The issue my colleagues and I see over and over again is the effort it takes some researchers to pierce the edge and be innovative.

Coming to the aid of these struggling researchers is Moore’s Law.  The reason why Moore’s Law has persisted is that we use tools we’ve already created to come up with new tools. With each advancement, it gets easier to create and we end up with exponential increases in almost every area related to computation - including robotics, mechatronics, medical devices, and unmanned systems.

The biggest dichotomy in research progress now is between those who are adopting already developed platforms and components versus those who do it all from scratch... from bolts to code.  The latter still happens.  

In any physical research platform there are four key components - plant, power, data acquisition, and software.  Depending on the focus of research, we can accelerate research by adopting as many pre-built components as possible:

Software - open architecture that can allow for quick development of controllers and extend already existing code

Data Acquisition - reliable board spec’d to suit the application.  Building data acquisition (unless that’s the research focus) from scratch is usually a huge time sink for researchers.

Power - reliable power supply is very basic - very few researchers build power supplies from scratch now.

Plant - this is usually the heart of the research. Unless there are already existing plants that can be used for research, this is where to focus should be.

There has been an explosion of creativity in the DIY community because of the increasing ease and falling cost of the software, power, and data acquisition. This creativity is only starting to spill over into engineering research.  The mindset of throwing a limitless number of grad students at a problem is slowly evaporating and being replaced with finding the tools that will get researchers to publishable content more quickly.

As the price of these tools drops and they get easier to use, the challenge of innovating will be more on coming up with new earth-shattering ideas.