Productizing The Internet of Things

Recently, I wrote an article for LinkedIn Pulse. Here is that article in its entirety.

CES 2015 is happening again as I write this, and the headlines are chock full of promising new gadgetry. Smart home hubs, connected appliances, intelligent entertainment devices, autonomous cooking machines, robotics, drone-toy-things, and voice commanded apps — and lots of them — are all taking center stage, promising a not-so-distant future rich with sentient computing.

The link that connects all of these products is the notion that product users want to spend less time interacting with products and more time doing fun stuff or being productive, which ironically may actually also revolve around interacting with more products. With product automation, device interoperability, and the promise of miniature computers that can remember your habits, one may be inclined to wonder what is changing in technology that is allowing all of this new innovation, and where is it headed? Well, some of the things that make all of this possible are data, hardware, and ecosystems. In other words, you’ve got products, the tools that build those products, people’s experiences with those products, and the marketplaces that sell those products. The first thing that has begun evolving at a warp speed in the last couple years is the tools for building products.

A major aspect of the modern super-developer’s toolkit that often doesn’t get enough credit is the way data is stored and analyzed. Data storage and analysis in its most fundamental form is a matter of disk space and query power. However, now that scalable data layers are built and consumed in the cloud, disk space is just a concept, and extracting knowledge from data is essentially some form of Map Reduction, which is grabbing subsets of relevant data and asking questions against those subsets until knew knowledge is formed. Additionally, computing infrastructure as a service is so easy to get now that product development life cycles are shortening at a rapid rate, but thats another topic for another day. In modern IoT, a whole lot of data is getting ingested at a massive rate, which isn’t new, but what is new is now those data layers can actively ask questions over large datasets in nearly real time. Hadoop is a big buzz word these days because it aims to do exactly that: ingest a ton of data in a short time and answer questions quickly. By utilizing these tools more, little products can have big brains, and thus, we get something to the effect of pseudo-sentience. That often translates into products that are helpful. Like, an air conditioner that knows to turn on because you just got home from work.

Another expanding aspect of Internet of Things products that is going to change everything is interoperability. Right now, most grassroots developers develop apps to interact with APIs in the cloud, which dispatches and receives commands not only to that device, but all other devices it interacts with. The effect is that if you perform an action on device A, you can see that command happened from device B. That’s the Facebook conversation effect. Think of it as a constant dialog that is being brokered from above, but is visually rendered in front of your eyes. It gives the impression that those two devices are actually talking to each other. In order for the IoT to truly be catapulted into the next era of interoperability, devices will need to improve their intercommunication at a hardware level. However, when it comes to hardware development, companies want standards in place to base their hardware choices on so their products don’t end up being derelict in a year after coming to market.

Additionally, its not just enough for those products to play nicely today, they must also play nicely tomorrow. We all must assume that these products will reach the maximum audience when they strike a perfect balance and achieve harmony with the other products users feel comfortable using. This is the essence of devices and app ecosystems. Users can and should be able to pick and choose what they want to use for every little action they want to perform. This is one of the reasons why smart home hubs have been a difficult problem to solve: how do you invent a central computer for your home that brokers all actions and information between your appliances and devices without requiring that every device in your home be part of that same product line from the same vendor? Nobody would buy that, and developers know that.

Ultimately, what is the most encouraging about what we’re seeing this year is that developers at both the grass roots level and the enterprise level seem to have understood that a certain level of openness will be required to achieve harmony between devices. Once those problems are solved, the true product development will start — after all of the version 1’s are completed, now these companies will fine-tune their products to be maximally useful to people. It is only then that these products’ user experience will exude a sense of unpretentious omniscience.

The only question is, once these products become perfectly good at doing everything for you, what will we do then? All of this is supposed to improve productivity and make way for more leisure, in theory, but it may end up just causing us to become perfect consumers — buying things so we can buy other things that automatically buy other things. Or we could all just go outside and take a walk.

Comments are closed.