Exploring software, hardware, everywhere

A Twitter Q&A follow-up to my conversation with Tim O'Reilly.

Last week, Tim O’Reilly and I sat down in San Francisco and had a conversation about the collision of hardware and software. The fact that digital entrepreneurs see hardware as part of their available palette now is really interesting, as is the way many companies with traditional manufacturing roots are seeing digitization and software as key parts of their businesses in the near future. Software plus more malleable hardware is like a whole new medium for building products and services. We really are on the cusp of interesting times.

As our time wound down, questions were still coming in via Twitter. Since we couldn’t get to all of them during the time allotted, I thought I’d try to respond to a few more of them here.

That is the business model we see developing in wearables and consumer gadgets like the Nest, but I think there are a lot of other possibilities. First off, as Jon Bruner and Mike Loukides highlight in their report Building a Solid World, it’s now possible to build “Everything as a Service.” An information layer built on top of hardware devices and the networks to share it lets us build services on top of locally present but remotely operated hardware (e.g. remotely operated gas power turbines), or to purchase assets fractionally rather than own the whole thing (e.g. car sharing). Time sharing is coming to everything, not just beach-front condos.

Even within the wearables area, if we consider a broader definition to include medical devices (or medical coaching use of fitness devices) we’ll see hardware devices bundled in as components of coaching or long-term care relationships.

I’ll take a stab at this for now, and Tim can weigh in later in the comments if he’d like. Certainly, we see the beginnings of this kind of thing in areas like home automation. The problem right now, as we mentioned in our conversation, is that many of the standards that will let sensors, actuators, and devices from different vendors work together are still immature or non-existent.

We are also seeing this in industrial settings as well, where companies like Enbala are instrumenting industrial electrical loads and then remotely controlling them at millisecond time scales based on signals from the power company to smooth and contain peak demands. The result is real savings in overall infrastructure cost and is an example of how information can reduce the need for infrastructure.


I think there is a tangential issue as well, which is that a lot of startups are running around with a new component as their starting point and are trying to figure out something useful to do with it, instead of starting from a need and building from there. “Ooh, galvanic skin measurement! Add it to a watch with an accelerometer!”


I’m not being facetious — I really think the point here is that hardware + software is qualitatively different than either of them separately. It’s a gestalt where the sum of the parts is greater than the whole.

I’m not sure that’s as big of a difference as you imagine. A lot of software (outside of the web space) has had to deal with regulatory and other hurdles. I think this is part and parcel to the fact that while hardware is becoming more malleable, it’s not going to suddenly have no hurdles at all. I don’t expect the cost of hardware startups to be as low as software startups any time soon, especially in areas like medical devices. However, they will be much more approachable than they would have been before.

Yes. Hardware is becoming more accessible to a whole range of entrepreneurs because of the developing generative ecosystem. Access to funding (Kickstarter), open source hardware and prototyping tools, frictionless manufacturing, etc., will make new industries possible (e.g. drone agriculture) and others ripe for disruption (e.g. satellite imaging)

I suspect he would have thought it was messy, and like the Homebrew Computer Club before it, worth watching closely for the moment when a company like Apple could come in and own a big piece of it.

Good question. One of the things we look for at O’Reilly are new skills coming together. In the IoT world software, embedded software, industrial design, and hardware design (both mechanical and electrical aspects) are all merging. I think that some of this will manifest in the need for new skills. For example, many software developers will pick up embedded skills as a necessary part of the toolkit, and more industrial design programs will follow the lead of schools like Georgia Tech that have been building software prototyping labs into their programs. But we’ll also see startups building around different kinds of teams, where industrial designers have a much more prominent role than we’re used to seeing in Silicon Valley to date. This will be a mix of individual and organizational learning and adaption.

Thanks for your questions! Let’s keep this conversation going — please submit additional questions in the comments, contact us via email, or tweet to @Radar.

tags: , , , ,