James Steidl - Fotolia
Supply chain software is evolving. It's increasingly moving to the cloud, and it's gaining intelligence with machine learning, internet of things (IoT) integration and big data analytics. The supply chain software market is evolving, as well, as innovative new companies enter a field that had long been dominated by three major vendors.
In this interview, conducted at the IDC Directions conference held in Boston in March, Simon Ellis, program vice president of IDC Manufacturing Insights, discusses the current state of the supply chain software market, and where it may be heading in the near future.
Who are the major players in the supply chain software space?
Simon Ellis: In the supply chain space, the big players are the 'big three' -- SAP, Oracle and JDA.
Then there's a host of notable vendors who are a bit smaller. Infor is at that level, Logility, Kinaxis, and E2open is another interesting one in that space now. They're not necessarily less capable, but are certainly smaller.
You will still see large manufacturers that use Infor or Logility, but they also appeal down-market. So they appeal to smaller to midsize companies.
Has the supply chain space had any cloud players or disruptive vendors, like a Salesforce, for example?
Ellis: Disruptive cloud-based competition has not tended to be true in the supply chain. We're talking about integrated business planning, and that has not tended to be true.
The disrupter in the space has always tended to be viewed as Kinaxis. They started out on supply chain responsiveness, more on the supply side, and they have always been on the cloud. They have a single tool, and they have augmented that tool to add in new capabilities: Demand planning, demand sensing S&OP, [and] control tower are now part of this single rapid response tool. Its heritage still is response management, which is something that appeals to discrete manufacturing and high tech, particularly, and the majority of Kinaxis customers are in the high-tech space. But [they have] generally been viewed as the disruptor in the marketplace.
I think SAP certainly views their sales growth of SAP IBP [Integrated Business Planning] as comparing to Kinaxis' [for] being the benchmark in the space.
Do any cloud-only players focus specifically on supply chain?
Ellis: There have been plenty of cloud-based ERPs -- Plex [Systems] and others -- but in the supply chain space, it's fairly early in the game, still.
Even NetSuite, which is a cloud-native application, has fairly modest supply chain capabilities, and I certainly wouldn't rate them. They're a very interesting company, and they may well do great things, but from a supply chain-only perspective, they have fairly modest capabilities.
Why has supply chain been slower to move to the cloud?
Ellis: There are a couple of reasons. I think that many companies hold their supply chain capabilities close. And supply chain tends to be a more inward-looking capability, even though there's some collaboration on the ends.
The other thing is we've had cloud-based execution tools for a while -- Transplace has been in the cloud for a while -- but I think transportation and fulfillment [have] been somewhat more easily adapted to the cloud.
Planning is a bit more tricky because it connects to so many things, and the ways in which companies manage their forecasts really is central to the differential value that they provide. It's just not something that they really want to expose, or have historically wanted to expose to the cloud.
But that's changing, and the resistance to cloud is almost completely gone. It was 'It will be a cold day in hell before cloud …' a few years ago, but now it's 'If you don't have a cloud option, we're not going to consider you.' So, the sentiment has shifted dramatically, and I think it was just more that the perception was that the inherent collaborative benefits of cloud played more to the strengths of sales, marketing, ERP, those kinds of things.
So that's why those outward-facing interfaces, like SRM [supplier relationship management] or CRM [customer relationship management], went early, and why you have Salesforce being as big as they are. But those of us who have been long-time students of the supply chain have felt that cloud was always simply a matter of time.
Where does IoT fit in?
Ellis: [SAP has] talked a lot about IoT, and the linkage of IoT with the SAP Leonardo platform.
At the end of the day, IoT is about getting more and better data. The sensors are material only from the perspective that they're capturing new data, so it's ultimately an analytics problem, at least in the near term. You can argue that, in the long term, sensor-enabled things and smart things will interact themselves, and [will] make decisions and do things without human intervention, but we're not there yet.
But you think that's where it's going?
Ellis: I do, absolutely. I think IoT and smart things [will interact] to make decisions within certain parameters of acceptance. If there are standard deviations of exception, then you'll want human intervention, but if there's not, then these things can just interact and essentially run it for you. That's the ultimate goal, and where we'll get to.
Ultimately, it's about how we leverage these different and better data sets to improve forecasts, to better understand our forecasting, and how demand and supply interact. Then, on the execution or fulfillment side, [it can enable] knowing where things are or having a better handle on delivery dates and where there are going to be exceptions.
Like if a shipment is lost at sea, for example?
Ellis: Or even trucks on the road -- it's supposed to be unloading now, but we show the truck at Louie's Luncheonette 10 miles away. It's not going to unload for a while, so they can dynamically adjust the schedules. If you've got a bunch of trucks waiting, and you now know this truck's not going to be here for 15 minutes, you can slot that in. So you can do a lot in the fulfillment space using sensors.
IoT is interesting in the context of supply chain planning, integrated business planning and supply chain execution, even if the long-term vision for smart things and smart systems is more autonomous and AI-driven, but we're not there yet.
It doesn't matter if the system is cloud or on-premises when including IoT?
Ellis: No, it doesn't. Of course, the sensors are, by definition, on-premises because they're on things, but the data will go into the cloud. I don't think you'd bother to store it locally because then you wouldn't have access to it all at the same time in the same place. So the data repositories will be centralized in the cloud.
There's just too much data, and the trouble with sensor data is that it's mostly redundant. If you think about sensors on something like a truck, it says 'I'm here' continually, and you may get a thousand 'I'm here' messages before you finally get an 'I've moved.' You don't need to [see] the 999 'I'm here' messages, you just need to know the first 'I'm here.' Then you need to know the 'I've moved.' So the system has to be intelligent enough to know the ones that matter and ignore the rest. Maybe they still get transmitted to the cloud, but it's just that the database, the analytics knows that I don't care about that. It's the change status that I care about.
Are there other examples of specific IoT applications that can improve the supply chain?
Ellis: One great example of the interaction of IoT and the supply chain is the ice cream business.
What's worse than having melted ice cream? It's having ice cream that melted, then refroze. But you don't know [that] because then it looks fine, but it tastes terrible.
So using IoT and sophisticated sensor capabilities to monitor the transportation and fulfillment processes such that, if the sensor goes below a certain level, the system can tell you that the quality of the product may now be suspect, and that's a huge benefit. It's the same for meats and so on, where you can get salmonella if they're not stored properly. That's not usually because of the fulfillment process, it's usually because the stuff got dosed at the factory, but it certainly could be in there.
You can also have sensors that detect potential tampering. There's all ways of bringing IoT into elements of the supply chain.
Here are some important keys to effective materials planning
Get a handbook on supply chain analytics