Interoperability is essential to capture the full benefits of the Industrial IoT. Using the PackML standard to monitor machine states could be the solution. In this video ei3’s CEO, Spencer Cramer, shares his insight about the role PackML can play within the Internet of Things. Recorded live from the Pack Expo Innovation Stage on November 8, 2016.
At ei3, I’ve been building industrial internet solutions for about 15 years and my topic is how business value can be created by putting PackML into the cloud. But before that I want to share some big numbers with you. In arecent report from Mckinsey Global Institute they say that if policymakers and businesses get it right, linking the physical and digital worlds could create up to 11 trillion dollars per year of economic value by 2025. Wow! 11 trillion that’s a huge number. But where’s all this value coming from? Well in the same report Mackenzie shows that the setting that creates the most economic value from the Internet of Things is factories. It’s not smart street lights in cities. It’s not refrigerators that order their own milk. It’s not wearable T-shirts that measure your heart rate. It’s factories.
What that means is that we are here, sitting on top of the largest setting of Economic Opportunity of the Internet of Things. That’s pretty exciting, but also in the same report Mackenzie has a little bit of a warning. They say that interoperability between IOT systems is critical and it’s required to unleash forty percent of that economic value on average and up to sixty percent in some settings. So that’s where PackML comes in. Interoperability is loosely defined as the ability for computing systems to exchange and understand each other’s data. PackML gives us a way to exchange and understand the data from different machines. It plays a key role with the Internet of Things.
In my topic today I will be talking about how PackML’s common description of machine states and modes transmitted with pack tags gives cloud solutions in the IOT the ability to compare apples for apples,do benchmarking, and big data analytics. I’m also going to share the fact that PackML is not just an Internet of Things solution as you saw from the previous two presentations, but also that PackML has a very strong ROI. There’s plenty of good reasons to use PackML inside the walls of your factory. Additionally, an added bonus is that you’re future proofing yourself because PackML has a valuable cloud proposition which allows economic benefits to come from big data. Being able to benchmark machines and from those bench marks understand which machines run well and which don’t, you can define predictive models from prescriptive services so that uptime and availability can be increased which boosts productivity.
PackML is great for integrating machines. So if you want to put a line together you can use PackML to integrate the states and modes of the machines by exchanging pack tags. PackML is a standard defines those pack tags. Also PackML works really great with OPC-UA. So If you have OPC-UA your defining the context and the structure of the data warehouse that moves the pack tags around and the OPC-UA will rest on top of yet another standard.
Ethernet defines a network enabling digital communications to occur between the Machine controllers so those three standards work together to enable the line integration on your factory floor. Additionally, those same three standards can bring data from the line up into the cloud with existing technology. It’s possible to use Ethernet, OPC-UA, and pack tags to bring data outside the walls of the factory, through secure networking methods, and out onto the internet. Data collectors can operate and gather the data and then can be put into private and public clouds where data from many machines, coming from many different customers, in many countries on different continents, can all be brought together. With that big data, sophisticated applications can do the analytics where data is converted into actionable information. This can then be disseminated to the authorized users in the form of webpage dashboard reports, mobile apps, and restful API’s to exchange the data with the right people, with the right systems, at the right time. So let’s look at an example, let’s say you have a machine that moves packages around, you’ve implemented it with PackML and with PackML through the pack tags you are able to understand how much time it spends on executing its operations. From that you can create an OEE profile. You can then understand how the machine is running and how it ran in the past. That’s very valuable because this is the data you need to manage the machine effectively.
Also if you have three machines on the same plant floor it’s great to be able to compare them. Because you want to measure to manage using data to make decisions. By knowing if a machine is stopped, or if it’s running poorly you can then do a Pareto analysis and with the Pareto analysis, as six sigma practitioners know, identify the best return on investment. If you find the largest reason for unplanned downtime and you spend your time and energy reducing it, that’s going to probably give you the strongest return on investment. So this works great and it’s already being done by many organizations inside the walls of their companies. Now imagine if you were to do this around the world, if you could have that performance data from machines running for different customers on different continents. Well, the OEM who builds these machines could have the ability to spot isolated instances at remote and different facilities. At these facilities there might be some common thing that goes wrong with the machine, and the OEM can then learn from those isolated instances and from those learnings can develop a proactive support. Where parts can be dispatched, or service technicians can connect through secure tunnels and upgrade the software.
With a solution like this now everybody who owns the machines from this company, even if they haven’t had a downtime, can enjoy the benefits of a predictive support solution. When they share data about their machine operation and exchange, they have a proactive service being delivered by the OEM. And of course with the OEM, whether it’s dozens or hundreds of machines at the same time out in the field you can very quickly get thousands of machine years of experience, and these thousands of years of experience can lead to very accurate predictive models. These predictive models can then be used to do predictive maintenance solutions. So with us, with this type of support, an OEM can offer his customers the ability to understand where that customers machine is in with respect to its peers. You can compare it with its peers in the same country or, all around the world. Having this information gives one the ability to develop predictive and proactive support services. PackML, provides interoperability and that interoperability is really good inside factories. It has a clear ROI and it shortens the startup. Also it simplifies the Machine operations making it easier to train and it makes the code modular easier to deploy.
All of these are really good reasons to put PackML in the cloud. Furthermore, PackML in the cloud has great advantages which include allowing you the ability to standardize your data and, with that standardized data then benchmark machines. From the benchmarking of the machines you can understand which ones don’t run well. From that understanding you can develop predictive models. Modeling behavior can lead to better predictive maintenance systems and, proactive support services. I see PackML in the cloud enabling all of us to be on our way to enjoying the benefits of the IOT.
At the beginning of this presentation I mentioned the Mackenzie IoT number, 11 trillion dollars per year or almost four trillion per year of economic value in factories. I’m going to challenge you all, what’s your share going to be? Because if there’s trillions of dollars at stake, there’s billions here in the packaging industry to be had, and millions for the companies that get it right. If you’re interested in capturing this value, I have two thoughts to share with you. One is use packML and if you want to use PackML the best place to start is OMAC. OMAC has lots of tools and resources, as Brian shared, that can help you get started quickly and effectively using packML. The stakes are high and, the solutions are complex. Two, if you want to put a system online on your equipment or in your factory I recommend finding a partner company. A company that can help you take everything from the Machine through to the cloud with private secure networks. Data warehousing applications and alerts can then bring you the information you need. With those two strategies you can save time, because you can take advantage of systems that have already been developed and, you can save money. Also you can reduce your risk.
Thank you very much for your time. Also, I want to say I’m really glad ei3 is a part of OMAC. OMAC does a great job bringing the right companies together, both machine builders machine owners and technology providers, to have a very lively dialogue. In that lively dialogue we can develop standards like PackML that enable you to have benefits for your machines today and also develop standards to benefit you in the future.