Paper written for the Manchester Developmental Panel on Innovation (New Economy and The Northern Way, 23-24 Feburary 2011)
Question – How does infrastructure become more innovative?
As cities become programmable, sentient, so the interface between physical and digital systems becomes increasingly significant.
How does infrastructure become more innovative? Deployed infrastructure needs to be tested, reliable, generic. But there is space for innovation in its development, in the creation of the wider ecosystem, and the development of services. We can create infrastructure that enables innovation. Stimulate a mashup culture so that people can build new services.
One way to stimulate innovation is to make infrastructure more open ended, open source. This needs to be done in a controlled, managed way. You cannot give an open API to critical infrastructure. There are real challenges, if you open systems up they become more open to abuse, they can become more invasive, and privacy and control can be challenged.
The lifecycle of big, physical infrastructure can be 15 years, and soft, digital infrastructure only 40 weeks. Digital systems need long-term sustainability, and big infrastructure needs to be more responsive so that it can evolve and adapt to changing circumstances.
A technology only comes to life once it is in use, and many applications are not foreseen, as in the example of SMS. Co-design and open innovation need to be a part of infrastructure development. Deploy prototypes in living labs, let users shape it, give access to artists, be open to unexpected, surprising outcomes.
ROI is not always the best metric when seeking innovation, as it encourages short-term gains. There needs to be space for “creative misuse”: make infrastructure free to use for non-commercial use, let researchers, students, artists, even early-stage bootstrapping startups use the infrastructure for free under reasonable constraints, including the transition to being a paying customer if they are successful. Playful experimentation can support innovation and make it more useable and friendly, overcoming barriers to adoption.
Building infrastructure as a single, monolithic entity can only done by the large technology companies. One alternative is confederated projects with many small contributors. These require an environment in which people work together, and a supporting infrastructure to track contributions, split rewards.
The traditional way of doing infrastructure is linear, create the technology and a business case, and roll it out. There is increasing awareness this does not work unless you also support the development of the wider ecosystem and engage communities of users.
DataGM is an example of an infrastructure designed to enable innovation, it will succeed when people innovate. Another is the PD-Net project at Lancaster University which approaches networked displays as an open platform instead of a closed technology, and is developing an ‘app store’ for public displays.
Drew Hemment, 24 Feb 2011
FutureEverything & Lancaster University
Thanks to Julian Tait, Adam Lindsay, Adrian Friday
