Is Edge Computing Analytics a real Internet of Things (IoT) trend for 2019 or is it more smoke from analysts and large technology vendors? Knowing that hardware manufacturers like Cisco, HPE, Dell and more are building specific infrastructure for the edge designed to be more physically rugged and secure, we need to believe there will be a lot of money at the Edge.
During the last months I have been aware of the difficulties of develop IoT Edge Computing Machine Learning solutions for the manufacturing industry.
How many companies have really combined knowledge of automation (PLCs, SCADAS, HMIs), with electrical and mechanical knowledge and with a unique Edge Computing distributed system to develop Machine Learning Applications to develop Industrial Internet of Things (IIoT) applications for a specific industry? Very few, right?
How many companies have installed IoT nodes in machining systems able to extract, transport store, transform and deliver valuable insights? Very few, right?
During my research and analysis of competitors, I understood that not all OEE, Condition Monitoring, Energy Efficiency, Quality Control or Predictive Maintenance applications in the IIoT market are the same.
I also have noticed one more time how the fragmentation of Edge Computing solution vendors, the noise and expectations around Machine Learning, the integration with IoT platforms vendors and the lack of experts that analyse, and advice companies about IIoT strategy, ROI and business models are delaying the adoption of these solutions.
Here below, some personal observations that I believe would help accelerate implementation of edge computing and machine learning application in the Manufacturing industry.
Avoid OT Vendors Lock-In. We need machine data availability
The top PLC manufacturer’s list place Siemens at the top spot worldwide followed by other major PLC manufacturer’s like Rockwell (Allen-Bradley), Mitsubishi, Schneider, Omron, Bosch , Rexroth, Beckhoff, Fuji, Toshiba following.
When we are developing Edge Analytics Machine Learning applications we need to extract data from the PLCs of the manufacturers. We could think that it will be easy using OPC open protocol, but not all PLCs have an OPC Server. Different models of PLC support various forms of I/O, so we can read data from a PLC by connecting our IoT node to one of the interfaces supported by the customer model of PLC and configuring our software to match the settings supported by the device. In fact, we can find many ways to extract data from PLCs if the manufacturers will provide info how to do it. However, Siemens, and most of PLCs manufacturers do not allow “easily” to extract data to third parties, neither their own customers.
Not to mention thousands of old machines with old PLCs without available info from OT vendors and the fear of customers to allow access PLCs to third parties.
It is not a question of protocols, it is a question of vendor lock-in and data availability
It is not a question of protocols, it is a question of vendor lock-in. For years PLC´s vendors makes customer dependent on their products and services, unable to use another vendor without substantial switching costs. Now the Industry 4.0 / Industrial Internet is calling to open their doors. Customers must be brave and claim for openness and avoid lock-in for 20 years more if they want innovation in their plants.
Edge Computing and Machine Learning the last frontier to break between IT/OT
In my article “IT and OT, Friends or Foes in the Industrial Internet of Things?” I was optimistic about the convergence of information technology (IT) and operations technology (OT) because I believed that IT and OT convergence could drive Industrial Internet of Things adoption. I really thought the days of discussing the convergence of IT and OT was over. I was wrong. If you can visit manufacturing companies´ plant floors, you will see how much work still need to be done.
Edge Analytics is a key component in the integration of IT and OT and requires the combined knowledge of OT and IT work together. But the lack of skills in both areas and the impact in the operations and business makes difficult which department should lead the Edge Analytics projects.
I do not see manufacturing companies will merge their IT and OT departments in the short or medium term.
Manufacturing companies need a role with authority (Chief IIoT Officer or CIIoT) and resources with business, management and technical acumen to lead the strategy and projects that affect both areas like those related with Edge Computing Analytics.
To Cloud or not To Cloud – To obtain ROI at the Edge you do not need the Cloud
When I wrote in 2016 “Do not let the fog hide the clouds in the Internet of Things”, the hype around Edge Computing and Machine Learning started. There was a confusion about fog computing and edge computing and how this layer will impact in the IoT architecture, specially to Cloud workloads.
The IoT Cloud vendors were disappointed because they saw their potential clients had an excuse more to delay their decisions to implement industrial IoT solutions, especially in the manufacturing sector that was one of the main sectors to adopt IoT thanks to the numerous Industry 4.0 initiatives. Three years later, the three Top Cloud vendors (AWS, MCSFT, Google) offer IoT platforms and tools that combine Cloud and Edge application development, machine learning and analytics at the edge, governance and end to end security.
Siemens have launched MindSphere, an open, cloud-based IoT operating system based on the SAP HANA cloud platform. The good news is that Siemens cloud platform is characterized by three dimensions of openness.
To obtain ROI in an Edge Analytics / Machine Learning solution is not mandatory the Cloud, neither wait years for build an infrastructure to integrate the OT and IT world. Manufacturing companies can develop or deploy Edge Computing – Machine Learning applications quickly to monitor the health of their machines or to improve their asset maintenance or the quality control of their plant floor processes without the Cloud. And the ROI can be significant.
It is not about creating more islands or mushrooms in the IT/OT environments with Edge Computing / Machine Learning applications, but in implement projects that help companies in the manufacturing sector to improve their competitiveness
Connected Machines is the only way for new Business Models
Security probably is being the main challenge for adoption of IoT, not only in the manufacturing industry. Manufacturers have been reluctant to open their manufacturing facilities to the Internet because the danger of cyber-attacks.
But we are heading for an economy of platforms and services that need products and machines connected. Every factory should be in a position to tap into machine data remotely and make it available for IT/Cloud/Internet applications.
This requires every Edge Computing / Machine Learning system implemented to be built with the capability to share remotely data via open and secure protocols and standards like MTConnect and OPC-UA.
Having machines connected is the first step to make machines smarter and to build smarter factories.
Technology vendors together with Machines manufacturers are exploring new business models more attractive to the customers. Having machines connected is the first step to make machines smarter and to build smarter factories. At that time, we will see new business models to flourish, such as Remote Equipment Monitoring, Machine Equipment as a Service, Predictive Maintenance as a Service or Energy Efficiency as a Service.
The benefits of using Edge Computing / Machine Learning solutions are very attractive to manufacturers because allows minimize latency, conserve network bandwidth, operate reliably with quick decisions, collect and secure a wide range of data, and move data to the best place for processing with better analysis and insights of local data. The ROI in such IIoT Solutions is very attractive.
But they will never get these benefits if they do not step up and change your outdated attitude and follow the suggested recommendations and start soon their IIoT journey aimed at to provide manufactures tangible and innovative business value.
Thanks in advance for your Likes and Shares
Thoughts ? Comments ?