Business cases for M2M technology sometimes a 'tricky affair'Date: Nov 01, 2013
Sure, machine-to-machine technology -- otherwise known as the Internet of Things -- is cool stuff, but what does it really mean for business cases and new revenue streams the technology can enable for SAP customers?
"The business case is kind of tricky," said Benjamin Wesson, vice president of product management for machine-to-machine (M2M) technology at SAP.
"It's not a question of IT; it's not a question of lowering costs, per se. Sometimes it's a matter of opening up or even changing the way the business functions" in pursuit of new revenue streams, Wesson said.
In this fascinating conversation shot on the floor of the 2013 SAP TechEd convention, Wesson elaborates on building a business case for M2M, the central role that predictive analytics plays, and SAP's strategy for the Internet of Things, using technology like SAP HANA and complex event processing to make sense of the data.
The following is a transcript of the video posted above:
Talk to me about where we are in terms of overall marketplace, in terms of the Internet of Things, machine to machine and what's being enabled out there. What are companies doing?
Benjamin Wesson: It depends on the business model, but there are some key trends that we can identify. One is remote service monitoring. Whether we're talking about the compressor on a vending machine or an asset like a John Deere tractor, companies want to retrieve data from their assets in the field to get information, identify potentially anomalous conditions that might indicate a future failure, and then analyze that data. And sometimes you [have to] update the machines themselves in the way that they actually function in the real world.
Businesses are using these monitoring programs as an additional service. Sometimes, to open up a new revenue stream, it might be on top of a warranty. And sometimes it's to differentiate their brand. It changes the game. Imagine you're driving in your car and all of a sudden the oil light comes on. You've got to pick up the handset, you've got to call the mechanic, make an appointment, get the car in; it's a big hassle.
Now with the Internet of Things, your car company is calling you. BMW is calling you. They're saying, 'Hey your oil light is going to turn on in 72.2 miles. We've made an appointment for you. It's on your route, because we know where you're going based on your navigation.'
It's a totally different experience, and people are willing to pay for that service. It is not a luxury in some cases. It's a service. It's a question of it working or not working.
Some of those things we've seen in commercials. Let's talk about SAP's role in this. What is SAP's perspective? How do some of those technologies like HANA enable and drive that?
Wesson: Let me start by describing how this market is incredibly fragmented. A lot of small players in the market have put their focus on what [is] called connectivity, or installing an agent on one of these devices, be it a tractor, a wind turbine, whatever it is, and then establishing a communication channel between that agent and some service they're providing. Our position is that, over time, these machines are going to come off the assembly floor with connectivity already built into them.
Our focus is not on connectivity, it's not on lighting up a cellular connection. It's about making sense of the data. We do that using tools like event stream processor and HANA. What happens is that we're getting a real-time stream of data from these machines. We're analyzing that, and we're correlating various events.
Let's say we're getting a temperature from a tractor, and then we're correlating that with the external temperature that we're getting from a weather service feed, and other variables maybe -- the humidity, for example. And then we're able to identify if this is either a problem we should look at or business as usual, and the tractor can go on operating. Let's say we do identify a problem. The question is, OK, we've flagged an anomaly in the data. What do we do next? What does that mean? When is this tractor going to fail? That's where HANA comes into play. You could provide something even as simple as a linear regression.
You could say if this temperature continues to increase at this rate, it's going to hit this critical threshold within this time span. And, once it does, it's going to cause damage. Then the theme becomes, how do we kick off an escalation process to dispatch a technician, to drop-ship parts and fix the problem before it gets to a critical stage. You also questioned about trends, [and] that's a huge trend. Track and trace is a huge trend.
We have vendors that, like the compressor on a vending machine, need to keep things cold throughout its entire distribution cycle. This is called cold chain tracking. A key trend out there is we're trying to deliver by automating the reporting of movements of these assets and by identifying things like anomalous conditions in that compressor.
It seems like the business case is probably fairly clear for a lot of companies. What kind of interest do you see out there? What kinds of questions do they have?
Wesson: Actually the business case is kind of tricky, because it's a different buying center that SAP is used to selling it to. It's not a question of IT and it's not a question of lowering costs per se. Sometimes it's a question of opening up or even changing the way that the business functions. A lot of companies that manufacture assets are finding that as their industry becomes more and more commoditized, the difference between their products and their competitor's products is price. As soon as you get into that scenario, it's a race to the bottom.
So the question is, how do I maintain my margins by offering a service, or how do I shift some of selling these products to actually servicing them? We have one customer, and they are savvy enough to say, 'Not only will we provide this monitoring service for our products, we'll monitor competitor's products too. Do you also have those in your building? We'll make sure that those don't blow up either.'
And then when you get ready to replace one of those assets, you're going to be more likely to go with us. Business case is difficult because it's hard to quantify -- for example, how much revenue you drive from a new service display. What we're trying to do is make it easy for these people to take the first step.
We're going to market with co-innovation partners. We actually spend time on the ground with them, making sure that a solution works or satisfies an end-to-end use case that they have. We give them a pricing model that allows them to step into the Internet of Things peacefully. You don't have to spend a million dollars and you don't need high-priced consultants. You could start, for example, with vending machines, and including connectivity, we're talking like $1.99 per vending machine per month.
Those are the types of price points we're talking about. And because it's based on the number of devices within reason, then it's really cost affordable, both for a small business and for a big business. What we're doing is helping them with that business case by showing them real results. We encourage these businesses to actually try it. Take a random sample. Take half of the vending machines in the Mexico City airport. Hook them up to the internet. Monitor them using our services. And see if it adds value. If it does, roll out to the next 10,000 vending machines. Over time you'll start to see benefits.