Making predictions about future technology trends is not particularly hard. It's not exactly going out on a limb to say that IoT, machine learning and blockchain are going to be very important in the not too distant future. The important thing is trying to determine how these trends may affect enterprises, as well as what people need to think about as they plot their course for the digital future.
SearchSAP recently spoke with three SAP executives about five future technology trends -- augmented reality, blockchain, artificial intelligence (AI), robotics and contingent labor -- and some of their implications for enterprises and SAP users. The executives are Dan Wellers, SAP global lead for digital futures; Kai Goerlich, chief futurist, SAP Innovation Center Network; and Michael Rander, global research director, Future of Work at SAP.
1. Immersive technology and augmented reality
Dan Wellers: [The] biggest example of augmented reality is Pokémon Go, on the consumer side. Until recently, this technology couldn't deliver a believable experience, and, as recently as 2014, the market barely existed. There was certainly nothing on the enterprise side, and those applications are just starting to [be] built now.
On the retail side, this could be the ability to try on outfits in a virtual mirror that takes your sizes as you do it, or sitting in a new car model simulation to sense whether that's the right model. Lowes has its Holoroom in a couple of stores, where customers can design a simulated room [using] a tablet.[Dan Wellers, SAP global lead for digital futures] Dan WellersAnother big area is medicine, and using augmented reality to overlay diagnostic and treatment information over people's bodies, letting medical students practice complex procedures safely. Cleveland Clinic is actually turning MRIs into conventional 2D and 3D images that can be projected over the site of the procedure for training or diagnostic purposes.
Also, a company might use augmented reality to walk field service reps through repair processes or, in an IoT [internet of things] example, on a construction site [by] scanning an area and having it trigger data about real-time costs, supply inventories or planned versus actual spending.
All of these have data intensity as their core, and the message to our customers is that you should be thinking about this stuff now, and work with us to coinnovate, either by adding virtual or augmented reality into existing apps or thinking about new apps for new needs that are being discovered as we talk.
Wellers: It's important that we look at blockchain from an enterprise standpoint because it has huge potential, but there's a lot that's been written that's not really concrete. One [example] has to do with the need for scale -- can blockchain scale to the levels that enterprises really need?
A great example is shipping containers. A large container ship can hold around 18,000 containers, 50 pallets per container, 50 boxes per pallet. So, on one ship, you've got 45 million items to track, and if there's 20 million containers out there, the numbers get really big. So can blockchain scale to that level? We don't see that happening soon, but we know that somebody needs to look at it.
We're starting to look at how close we can move blockchain toward enterprise requirements -- what are the hurdles, what unknown things will we come up against? We believe that we can help based on our experience in the enterprise and [with] business process understanding and mission-critical systems. We're looking at approaching it from a full, end-to-end business process perspective; things like [increasing] process speed and transparency, improving supply chain efficiency and facilitating business networks and exchanges.
3. Artificial intelligence -- machine learning and deep learning
Kai Goerlich: The concept of AI is from the 1950s, and usually the connotation is more on smart, human-like robots. Now, basically there are two versions -- machine learning and deep learning -- and 2017 will probably be the year of the beginning of true AI. We use machine learning as the umbrella for both, but there is a subtle difference. Machine learning is learning by statistics and coming up with rules, while deep learning is self-learning with qualified data. At SAP, [we're betting] on the deep learning algorithms to hit business, and we think that those algorithms will change how we do business.[Kai Goerlich, chief futurist, SAP Innovation Center Network] Kai GoerlichIf you look into it, deep learning is very much like human training; you need good data and time so that the algorithm [can] come up with a solution. It could be that, if you feed the algorithm the right data about demand and supply and logistics, the algorithm might find a better routing, for example, or an improved supply chain. For the moment, deep learning has been used mostly, as you know, with AlphaGo, the algorithm that beat the Go world champion, but that's a very narrow frame; it's a board game.
We are now using the same algorithms for cars to drive by themselves or for image and speech recognition, but these are still limited data sets. If you look at a complete supply chain, that's significantly bigger, and with every additional parameter, the information base gets exponentially bigger.
We are at the early stage, and we will probably start to find [opportunities in] the highly repetitive stuff in the supply chains, like seeing if certain items have been shipped or not or if they have been properly taxed; something where such an algorithm would help.
4. The emergence of cyborgs or robotics
Goerlich: There's an underground cyborg scene in many big cities, and humans are testing cyborgs. There are some medical cases that make it into the news, and artists are using that stuff already where they try to bridge or hook the brain to the computer or have implants. It's underground, but with the medical use cases, it will be more prominent, and we will see that in some niche areas to start. It's a big testing area for some extreme or fringe use cases in robotics and algorithms; it's ideal for machine learning and algorithms because they can learn on very good data sets.
The future is difficult to predict, but my bet is on human-machine cooperation, or maybe even coevolution, so that we merge with what we now call machines. The other scenario is that we have very smart machines, like what I call the "Terminator Scenario," where they will be much smarter than we are and turn out to be very clever, but not merged with humans. So humans stay humans and machines stay machines. It's pretty appealing to some people, so I'm pretty sure that we will test it beyond medical applications.
5. Automation and the gig economy
Michael Rander: Automation will affect the future of work and how the workplace is going to look, but it's part of a bigger picture. It comes with all of the different advances in technology, so we look at it as the whole digitization of the economy and how that's going to affect the workforce.
Automation will move a number of jobs in a certain direction, and they're going to take the easy to replace jobs first; the manual ones or the ones that can be replicated the easiest. Over time, we'll see that it will affect certain groups; for example, taxi drivers, where automated cars could essentially take over that whole workforce.[Michael Rander, SAP global research director, Future of Work] Michael RanderThe problem we have then is, what to do when we have a whole group of people who find themselves without any training to go and take other jobs? So it becomes a larger societal problem. It's something that we feel we need to dig into from an SAP perspective, but also from [a] collaboration across companies perspective. We want to engage other companies, we want to work across boundaries to look at this because it becomes a workforce problem and becomes a societal problem. So it's something we need to address or we're going to have much larger problems in the workforce.
This means we need to look at training, at the hiring processes, at what it means to have a contingent workforce. How do you manage a contingent workforce and make them be a part of what you do as a company? It's estimated that, basically, every company will see contingent workforce as an essential part of their business.
So you can't just look at this and say you'll find a solution when you get there; it's an issue right now that you need to deal with. Looking just a few years ahead, you'll see that trend increase, and people that aren't getting ready now will fall behind.
We did a study where we were looking at the digital readiness of companies, and even though a lot of these companies say that digital transformation is an important part of what they do, very few are actually ready for it. People are realizing that it's happening, but they're not taking the steps to address it, which is a major problem.
Tying the future technology trends together
Wellers: The real value in these doesn't come from them standalone, it comes from how they combine. How do things like 3D printing, IoT, blockchain, connected cars, smart cities and the sharing economy combine to create a logistics internet? How do nanotechnology, robotics, augmented reality and connected healthcare combine to revolutionize the way healthcare is delivered and how research is done? How will blockchain, artificial intelligence and IoT contribute to the smart logistics supply chain?