(This blog published on Sunday)

Tomorrow let’s take a moment to remember those who have fallen. Not just here in the US but also around the world. Those who paved the way for those of us who are free and those paving the way for those who wish to be free.

(To those who have fallen).

Moving this topic from my person al blog here to IASA.

Andersen’s law of innovation: The rate of innovation remains constant until there is a change in pressure applied or capabilities available.

Yesterday I published a few patterns of Innovation on my personal blog (above). I’ve also talked about some of the pressures and capabilities for innovation that exist. The thing about innovation that we often forget is that it is a component of what software architects need to be prepared for. From infrastructure (can we support the innovation with the systems we have in place) to the solution architect (does this move us closer to what we need to place to succeed as an organization) and the business architect (does this move us towards our overall business goals).

In the end innovation is about expanding patterns and building new views of what exists.

I’ve started (based on a conversation with some fellow architects) the concept of innovation patterns. There are more than the ones I am sharing today but this gets you in the door and thinking about what patterns impact innovation and how.

Linear innovation can be measured as an ongoing improvement that seldom goes beyond the expected result. For many years Moore’s law was in this category. Processor speeds will double every 18 months. In fact the reality was it worked for many years. The linear pattern has a beginning and an end. I suspect if we spent some serious time looking at that pattern we could even create a mathematical representation of the start and stop of linear innovation.

Applied innovation is taking innovations that were created in another technology area and applying them to a new area. (the discovery of penicillin fits this innovation pattern). Applied innovation can take existing linear innovation patterns and move them in a new direction. The applied innovation only applies to the initial innovation and may in the end create two distinct linear innovation flows.

The last pattern we talked about was exponential innovation. This innovation can occur from either the linear or applied innovation patterns and is best defined as innovation that explodes forward rapidly. At the end of the Moore’s law processor cycle there was an exponential explosion of innovation that occurred (where the reality of processors went from 1 core processor to multiple core processors and now 4, 6 or more). This exponential innovation wasn’t reflected in the overall increase as predicted by the law instead this was a massive shift in total processing power.

I would like to offer that innovation is a conversation that every software architect should be having every day. Can we do this better?


Scott Andersen

IASA Fellow.