Industrial Internet of Things, enables going beyond the 4 walls of a Plant, to Mobile Plant Supply Chain

The last two blogs on the “Cyber Physical and Operational Management Evolution” and “How do you Achieve Orchestration in Industrial Internet of Things without Managed Configurations and Standards?” have created a lot of activity in hits but also email discussions.

This is good to see as two years ago these subjects would have hardly moved the needle, and they real opportunities for leading companies to embrace to expand their capabilities beyond the four walls of their plant.

Again last week I was involved in a number of discussions around these topics and liked one of my South African collaborators discussing how we have all these rich applications and capabilities for the fixed process plants, WHY cannot we apply these same tools to “MOBILE PLANT”?  Now he was from mining and was launching into the extraction side of mining and how to optimize the asset utilization, but he really wanted to go beyond that “Operational Optimization”.

The targets are not about data, what I like he is putting real operational goals in place:

• Operational Processes optimization, understand operational times vs expected times and analysis of areas to improve
•  Asset utilization / optimization
•  Energy and Fuel optimization

As he put we have platforms in the plants that abstract equipment below, and model these equipment so we can record, track their operations, and then apply operational process improvements, and built in operational process rules for the fixed plant. Now taking this to mobile these same platforms could be used but now across mobile equipment, so now we must record geographical data as core as location is key when using fuel, doing operational routes, and time is of essence. But the Delay accounting applications of today could be applied to these mobile equipment and we could then move beyond that to embedding operational best practices and operational behavior in the devices, and equipment to guide the operations to work within the “operational windows” of optimized performance.  

The diagram below shows a more detailed chart of maturities I mentioned last week. The concept of smart, to optimized and autonomous can only come with inbuilt operational strategies and practices that enable the orchestration I talked about last week.

Source ARC (arcweb.com).

The key is most of the mining extraction/ mobile plant is isolated, and I would say siloed even when connected between applications.

Remember the Industrial Internet of things is not about data, it is about “actionable decisions” in the NOW, by either machines, applications or people, and this will require embedded operational strategies/ processes that coordinate the mobile equipment to align with the overall business strategies.

If you are looking in the plant / fixed process world to apply IOT and gain significant value, you should think again and “open the door” and look outside the plant to mobile plant, or mobile supply chain, and extend the richness of operational applications to these traditionally isolated equipment and processes! 

How do you Achieve Orchestration in Industrial Internet of Things without Managed Configurations and Standards?

Last week I was at mining conference and had a rare chance to sit back and listen to people’s thoughts on innovation, and the future. It was good to hear the topics of partnership are key to innovation, (relating to my blog of a month “Participation architecture and culture key to Innovation”).

As expected the “internet of things” came up a lot, in many contexts, like it did at the Dairy conference the week before. With this cam the usual many definitions of IOT and the impact it will have on the mining industry. I just wondered how many people really comprehend the value, and complexity that it brings?

One evening I was on call with France with a partner discussing smart cities and IOT and he made the interesting comment:

“The Internet of Things has moved beyond big data and analysis to how will we align the devices and people into an orchestrated operational strategy that achieves a repeatable agile outcomes.”

I sat back with a big smile as he had articulated the change I had been seeing. As decisions and data is nice but it must go from data, information, knowledge to wisdom where actions can be taken, no matter if that action is taken by a device, or human.

Then I saw this categories of maturity in the internet of things, I had seen something similar but in a week of much discussion on this topic I thought this one would do. It shows devices going from a data sources with intelligent data / I hope actually Information. Evolving to control of devices in orchestrated way, no matter if the control is in the thing or in cloud the things know how to work together in a coordinated strategy. Once you have all the things working together you can move to tuning their operational behavior and effectiveness. This seems simple but things require access to control strategies, and orchestrations that guide these things, now we talking 100s to 1000s of things in this coordinated community. Eventually you end up autonomy or semi autonomy “managed by exception”.

In another discussion with a large network hardware supplier we were discussing a mining extraction alignment solution that could be enabled by IOT unlike today. So we took a practical look at the application, and saw 10s of like machines and a few classes of machines. Then you look at the operational processes they executing and again see repetition, but we are now talking 1000s look at devices.  Yet we had a customer wanting achieve level 3 in the above model “Optimization”. I thought back to many industrial sites I have been on in the last few years where there are 10s of PLCs programmed with larger control strategies but programmed at different times and by different people (even if from the same vendor) and how customers were having a significant cost of ownership in evolving these strategies. This why organizations like OMAC and PACKML have come about defining standard control strategies for operations/ devices that could span vendors.

So I ended back at my conflict, as we move to landscape where we will have 1000s of devices often smaller than traditional PLCs but each with their own monitoring, or control strategies, and then high level strategies that enable the orchestration of these devices/ things into a an effective operational strategy.

I asked how are we going sustain and evolve these strategies without having an “Enterprise Standards Management Framework” that enable standards to built for an operation? These are then deployed over 100s of similar operations on different devices. Now we shifted to managed, agile and sustainable solution.  

The thought of 100s of people programming 1000s of devices and then trying tune and evolve these seems un practical, plus if we enable standards management the reuse of IP and rapid rollout is achieved, while leveraging the revolution to smart devices and lower cost devices that execute these strategies.

A food for thought!!!!!

Smart XXXX: What does it mean!!!

So often today you hear the word “smart” put on the front of a segment describing the transformational program encompassing many of the Internet of Things concepts.

Smart Cities, Smart Farms/ Agriculture, Smart Airports, Smart Plants, Smart Fields etc.

Are they different or do they all come down to a basic set of concepts, transformations that are applied to that industry to significantly shift the needle in operational efficiency?

 Fair question, and so often lately I am being asked what is the difference between IT/OT, IOT, and Smart xxx? So I thought it was worth a discussion, as I suspect there different interactions.

To me the discussion of “smart/intelligent” industrial it is all about achieving “operational Optimization/ Excellence”, to suite the required production at the most effective time, cost. This is a shift from time based production and managing the process to managing the production of product/service. Driving the optimized execution of work / actions on operational processes for that product/service delivery.

At the core it is about changing the way in which we manage and execute work tasks, either automated or actions with human intervention so that only required work is performed at the correct time.  

“Smart Strategies” are fundamentally different from current IoT, Big Data etc. thinking:

• The IoT, Big Data etc. Initiatives/trends can be characterized as offering the 5 “any’s” – any information, in any context, at any time, to any user, for any action
• “Smart” products and operations can be characterized as offering the 5 “right’s” – the right information, in the right context (operations situation), at the right time (which is often earlier than “real-time”), to the right users for the right actions (which are often preventative and at best prescriptive).

All fundamental on the journey towards “operational excellence.”

That said “Smart Strategies” will employ the services of IOT, and big data, but the key is “Smart” is about tightening the execution of an operation process relative to the current product delivery expectations. A key concept is that the Operational Process, (no matter if it is in a city, airport, or production line) understands:

•        What it is expected to deliver in characteristics of product or service, and when
•        It is “self-aware” of it’s condition and ability to deliver that product/ service, due to capability, materials and the situation it is in.
•        It is able to then request and interact with other process, applications, assets and people to gain the required actions needed to succeed and when. 

This is a transformation from just understanding it is taking control of the process, as opposed to time schedule actions.

Information Driven Operational/Process Excellence Set Drive Next Wave in Mining but with a Twist

As I toured a number of the leading mining companies this week, the conversation showed a significant shift from last year from "greenfield" to “brownfield" discussions. Shifting from new plant implementation and speed to full production to how they draw the most efficiency from existing assets. The interesting twist was that the discussion of what was an existing asset:

1/ Fixed assets such as equipment

2/ plant ore assets

3/ mobile assets like trucks, digging equipment

4/ human assets, operators, maintenance and experts

So the strategy was how to tap existing information more than often locked within SCADA trend systems, and other data stores, it was key to extract this data and align these records into effective information. The driving forces are :

1/ minimal impact on the existing systems

2/ speed of delivery of the value

3/ expertise to understand and interpret the value

4/ predictive awareness, pattern recognition

The diagram below again resounded in the discussions.

The key for Industrial Analytic s is the trusted data, and just a historian will not achieve this, the model and validation must be done as close to the source.

The information needs delivery in many cases outside the automation landscape, often in the corporate networks. The key is to use the not APIs but make the connection through an SOA architecture. The service sits on the data source, with configuration, and data delivery built in, but key is low impact and effort.

This is not new, as the enterprise historian has been around for years, but the real difference is the need not just gather data, but to capture the data in a structure,  context, and validation of data that makes sure all stored data through resulting information is in trusted.

You are probably sitting there and saying nothing new! Fair, but the key was how are they going get this structured trusted data, that the concept was to do this as close to the source as possible, and then send through. This means the underlying systems do not change, minimize risk, maximize Lifecycle managed to enable evolution which will happen. Why is this not an IOT service, local and pushing vs polling, “self configuring” ?

Remembering the performance team of experts can be anywhere, and will probably virtual, where sharing, analysis and Modeling is done in offline mode looking for patterns.

As the discussions evolved the architecture evolved, and again the "cloud" came into play, why because the data size will grow, the users are everywhere, and the infrastructure of delivering is now there.

Why not?

The collaborative information, industrial analytics, is going to be foundational for the future of Gen Y teams of analysts experts from different locations and outside the companies.

Standby, as we see some of the optimization learnings from Oil and Gas come over into mining.