This article was originally published on CFO.com.

Data icebergs, drones, augmented reality, information on everything, the API economy — it should be a fascinating year.

After the Consumer Electronics Show (3,600 exhibitors with nearly 500 of them venture-backed; all three Las Vegas Conference Center halls plus 10 hotel conference center venues; 160,000 attendees) I thought I would sift through all of the themes and things I’ve seen lately and share some thoughts about what we might need to pay attention to in 2016. I’m no better than anyone else at predicting the future, so draw your own conclusions, but here’s where I’m starting from this year. —

The Information Revolution: Part III

Big data will get both bigger (because it can) and (to avoid being overwhelming for normal humans) smaller (more focused on what’s most important rather than what’s available). Someone will call the latter phenomenon “data icebergs.”

Information on Everything (IoE) will replace (or at least extend) the Internet of Things (IoT). Every “thing” in the IoT generates data. Without sensible standards for organizing and interpreting this data, IoE will be 99% noise. Even with standards (and without the development of effective filters and analytics) IoE will still be 90% noise.

Analytics (driven in part by improving situational awareness and IoE) will get both smarter (right more often) and less understandable (how on earth did we get that particular answer, even if it’s right?). The “democratization of information” will lead to rapid growth in misunderstanding and misuse by large numbers of (untrained) people, creating a whole new field for litigation.

Contextual simulation using prescriptive analytics (essentially “what if” analysis crossed with machine-generated sets of logically consistent but fake big data) will become fashionable among corporate strategists (and management consultants) but will turn out to be fairly useless (a fool with a tool is still a fool).

It will be hard to tell the difference between corporate “data lakes” and data landfills. Data pollution (in the data lake) will start to matter because deliberately or inadvertently corrupted data will affect decision making. Effective data governance will continue to confound almost everyone as the volume, variety, and velocity of data overwhelm our governance frameworks.

Machines (really the software that runs on them) will become steadily more “human literate” (because humans are finding it hard to become more technology literate). We will likely see this first in vehicles, then at home. The workplace will lag well behind.

Data protection and security generally will get (a lot) harder. Security through obscurity, which has been a common default tactic for a lot of businesses, will become nearly impossible. Finding and defending against the weakest link in the value chain or ecosystem mesh will become a critical security concept and a new service opportunity. An increasingly common condition for strategic partnership will be the ability to demonstrate adequate security capability — proof that you’re not the weakest link. Data breach insurance will become extremely expensive.

Augmented Reality (AR — overlaying data on the visual field to help understand what you’re looking at) and virtual reality (VR — avoiding looking at reality altogether) will be big topics, but will rapidly come to resemble digital graffiti rather than a useful evolution in information display. Outside of hardcore gaming, AR will win over VR at least for a while. AR hacks will be some combination of amusing, irritating, disgusting, and eventually illegal, which will do nothing to stop them.

“Information as a service” will be a hot concept in the enterprise, but IT and the business will remain separated by seemingly un-crossable chasms of culture, habits, history, and organizational inertia.

The Year of Software-Defined Everything: First of Many …

“Convergence” (architectural in technology and commercial through M&A) in enterprise technologies will continue. This is an interesting reversal of the technology sprawl of the past 30 years, during which the essential elements of business technology (compute, storage, and network) were separated and siloed. The result has become too hard (and expensive) to manage, so we are putting it all back into a set of “hyper-converged” super-appliances — betting that the technology vendors can solve the management and optimization challenges for us. Interesting idea. There are some early success stories. The jury remains out. Better not call the results “mainframes.” We are also seeing a convergence between the other great divide inside of enterprise IT: development and operations. The resulting “DevOps” actually seems to work.

Effective data governance will continue to confound almost everyone as the volume, variety, and velocity of data overwhelm our governance frameworks.

If everything is going to be “software defined,” however, the quality and reliability of the software we depend on better get a lot better than it has historically been. Hardware is actually very reliable — much more reliable than the software that runs on it. Now the most reliable elements are being relegated to commodity roles and the least reliable are being promoted to a critical role. Oh, well.

You’ll also hear a lot about the “API economy.” Opening up your transaction systems and data stores (soon to be lakes) so that you can collaborate better with everyone in your business ecosystem has a lot of positive possibilities for productivity and efficiency. Find whatever capability you need easily; reuse other people’s assets rather than duplicate them; pay (or not) for what you use; get paid (or not) for what you provide.

A lot of your business partners (and some of your competitors) will want to have this capability as a way to work better with and for you. Some companies will do it as a way to attract and build an efficient ecosystem. Some will resist, driven by security, loss of ownership, and erosion of asset value concerns. We are already seeing some of this last concerns in financial services around data aggregators. Expect some hard decisions over who owns what information and the right to do various things with that information. Plenty of real security issues here too.

The Year of the Drone: First of Several …

Drones as devices are already pretty well commoditized and prices are dropping rapidly. Drones as a platform will be big in selected verticals (but not in home delivery) in 2016. Winners in the drone platform wars will have an existing (and legitimate) FAA rule exemption and a real business model with added value over non-drone-based approaches. There won’t be many winners (the whole drone space looks a lot like the dot com mess redux). “Drones as a service” will probably be the best initial strategy. Value will come from what drones can sense and see and the analytics and information exhaust that the ability to fly predictable patterns at known times creates.

Like many technologies, drones will be misused. Because they are small and hard to spot (and can be made stealthy and hard to detect) they are potential commercial espionage platforms. Because they are limited in flight duration and prone to instability under some atmospheric conditions, they can drop out of the air suddenly — instant conversion to a bomb. You may have seen the video of the camera drone crashing just behind a downhill skier. Scary. Drone operation insurance underwriting is going to be interesting.

IOT at Home, Part 1

Lots of things in your home (appliances, environmental controls, lights, windows shades, door locks, security systems) can or will soon be smart — equipped with sensors, limited processing and data storage, and the ability to communicate with you via a user interface of some kind (probably an app) and with each other and the rest of the world via one or more communication protocols.

While devices are individually getting smarter, collectively they aren’t yet. IoT at home is early enough that it’s not at all clear what the best way to do this is. Architecture wars are inevitable — probably split between an “everything calls home” model where the communication is from the edge (a device or co-located group of independent devices) to the center and a “mesh” model in which most things happen locally and information only flows from the edge to the center when specific criteria are present.

Interestingly, both approaches will have connectivity challenges. Edge to center can rapidly overload the in-place network bandwidth currently available, especially when everything tries to call home at the same time (the flash mob phenomenon). Mesh approaches need more local capabilities to detect important local events and tend to eat up local wireless spectrum as the devices consult each other and share information.

Another challenge for the mesh model is the impact of false positive alerts and the resulting potential for first responder overload. If everything in your house knows to call 911 under some specific circumstances, the chances of a false positive go up substantially. Minor events can rapidly get magnified in significance unless there is intelligent local control and coordination. Too many false alarms are going to get you fined. Insurance premiums are going to get adjusted. The law of unintended consequences writ large.

My personal prediction favors the development of a hybrid model with a local mesh centered on a “smart hub” device that can call home. This is going to get very interesting.

The Year of Digital Health Care (Again)

I started my career in the health-care industry (way back in 1973) and I’ve been waiting for technology to trigger some meaningful changes ever since. I’m still expecting to wait. There have been lots of tactical improvements in the delivery of health care over the past 40 years, but the entire industry is still a mess — too many competing constituencies and interests, with every inefficiency in the system representing someone’s source of profit. Despite everything I’ve seen recently (much of it very clever), I don’t see this changing any time soon.

Finally, …

Don’t expect to see fully autonomous cars in 2016. Do expect to see a steady evolution of driver assistance and safety technologies. If the accident rate declines as these deploy, the case for full autonomy will be easier to make. The major auto OEMs have just about figured this out and are about to take over the evolutionary process from the technology firms that launched it. Not because they know the tech, but because they know the way (the vast majority of) people actually drive. Badly.

For all of the implicit cynicism in the preceding comments, I see 2016 as an important year on the road to a fully digital business and a digitally equipped consumer. The capabilities we need are there or almost there and most are almost usable. The value propositions are becoming clearer. The early adopters are far enough along that the general shape of the roadmap is becoming visible. We won’t get all the way there in the next year (or five) but we can certainly make effective progress.

About the Author
John Parkinson

John Parkinson is an Affiliate Partner at Waterstone. John brings extensive experience to the topics of technology strategy, architecture and execution having served in both senior operating and advisory roles.