Representative End User Clients
Representative Automation Clients
Representative Software Clients
The Honeywell Users Group EMEA Conference 2016 was held at the ‘world forum’ in The Hague, The Netherlands, from Oct 24th to 27th. One of the themes throughout the conference was IIoT by Honeywell. We researched how the solutions are architected, looked for early implementations, and deciphered what we believe is specific about the approach.
Common components we find in cross-sector IoT applications, and also in the process industries are virtualization, cloud computing, pervasive networking, big data, analytics and machine learning, smart devices, mobility and cyber security. IoT for industry needs to fulfill specific requirements and constrain the architecture:
Process safety reasons imply high-availability real-time control and fault-tolerance. Solutions must improve safety or in the worst case preserve current safety performance. Failures must allow continued safe operation, safe degraded operation modes, or safe shutdowns, in that order of priority. Safety takes priority over confidentiality.
Installations are capital intensive, are long-lived. These investments must be protected to guarantee satisfactory return on assets (ROA). This implies providing possibilities for upgrade and modernization for brownfield installations. Implementation of those must be done ‘on-process’. For profitability reasons, CAPEX and project cost should be kept to a minimum.
Other requirements define the capabilities of solutions: optimizing supply chains, reach higher production levels, help personnel improving or complementing skills, or making the use of skills more efficient. (For a more detail discussion of these needs, see our first blog in this series).
IIoT by Honeywell is a three-tier solution. The first tier is the classical process control with instruments, actuators, a DCS and a control room, with a few additions. First, instruments and control devices, including operator stations are cloud enabled or connected to the cloud via gateways.
Low cost sensors for purposes other than control, for example for monitoring and analysis purposes, can be added and connected via gateways. Collaboration of operations with maintenance, management and engineering on-site, with the help of advanced tools and analytics can further help improving performance. Honeywell proposes access to real-time and historical data with context based on the asset hierarchy, operational performance and business-oriented KPI visualization. Advanced solutions have extensive capabilities to discover abnormal behavior, analyze it, build analytics ranging from simple calculations to complex multi-variable model-based, and create run-time-analytics-based rules automating the early detection of abnormal behavior. For details on the corresponding products Uniformance PHD, Insight, Asset Sentinel and KPI, see “IIoT Analytics Platform Uniformance Suite”. Also operator stations are cloud-enabled, enabling remote operation.
The second tier corresponds to enterprise-wide collaboration. Where cloud-based analysis and expert support was an option in tier one, in tier two it is the only feasible solution. As all advanced solutions are cloud enabled this is technically possible. An important step to make this successful according to HPS is the ‘fusion’ of operational time series, alarm and event information plus other data such as maintenance records, originating from multiple sources in a globally accessible data store. Uniformance PHD can provide this functionality and thereby enable both engineering-science or data-science based analytics at global level.
The third tier extends collaboration of tier two beyond the enterprise boundaries to and ecosystem of third party equipment, engineering, operations or maintenance service providers.
The Cloud platform can be Honeywell’s Sentience platform or a platform of the users’ choice. The platform allows, amongst others, the development of apps by third parties, and which can be made available to users in the tier three scenario. (See also the presentation “Taking Advantage Of New Technologies”, presented at the User Group Conference). An example is HPS’ own connected services to assist users in maintaining performance of systems and applications, dubbed Benefits Guardianship Performa (BGP).
HPS reports the usage of remote operation of off-shore platforms in the Gulf of Mexico, the remote monitoring of draglines, shovels and trucks in a mining operation in South America, an enterprise-wide implementation in an multi-site refining and petrochemicals company, a predictive maintenance solution in a refinery on the Gulf coast in cooperation with Flowserve (tier three type), an inventory and work-in-progress optimizing involving multiple supply chain partners of a mid-size chemical operation and the optimization of a refinery in the US with help of expert support from UOP.
At the Conference, Ragnar Heksem of Lundin in Norway came to testify about the remote operations center his company installed on shore, to remotely control off-shore installations. Mr. Heksem mentioned the center has been connected using redundant optic fiber and is cyber secure. Up to 180 providers can log in securely. Mr. Heksem mentioned that safety, cost and resource efficiency were the drivers for the project. He said that it is more effective having operators and process experts working physically closely together. As process upsets are often du to human error, the company expects to reduce their frequency. The company aims at having less process upsets because of this. In many other respects remote operation is identical to local operation, Mr Heksem said.
From what we know about the solution today, it corresponds to the requirements for Industrial IoT. Built as an extension of existing control systems, it is independent of the function of providing safe and reliable control of critical operations. This protects the existing investments and keeps new investment to the necessary minimum, in particular since cloud solutions are fast and efficient to roll out.
The company has very well described the possible use cases and business processes that end users should take advantage of in designing IIoT implementation.
I believe that the fusion of time series, alarm and event data with unstructured data types enables the user up to take advantage of both engineering and data-science based analytics. HPS provides a very rich set of modeling approaches for analytics that we believe should be capable of creating solutions for virtually any situation, from simple to very complex. The user can choose from several model types, or use them in a cascaded way, complementing one another. The strength of first-principles engineering models is their capability to reliably make predictions in data-scarce regions. These can complement data driven models with faster and reliable convergence for on-line (closed-loop) usage. These two cloud solutions are complemented by real-time models running at the edge that can react on operational signals in split second. The visual analytics approach guiding the user in exploring and building monitoring rules, promises to be a significant source of engineering efficiency. More information can be found here. We believe advanced solutions are likely to enable faster detection, more appropriate decisions, and faster action on early and weak signals, thereby likely contributing to improving reliability and process performance.
Conclusions
HPS created a well-built and articulated IIoT architecture with a rich set of services and apps, built upon Honeywell’s Sentience platform. ARC will continue to follow product development and user applications and report about them.
Industrial versus generic IoT
Common components we find in cross-sector IoT applications, and also in the process industries are virtualization, cloud computing, pervasive networking, big data, analytics and machine learning, smart devices, mobility and cyber security. IoT for industry needs to fulfill specific requirements and constrain the architecture:
Process safety reasons imply high-availability real-time control and fault-tolerance. Solutions must improve safety or in the worst case preserve current safety performance. Failures must allow continued safe operation, safe degraded operation modes, or safe shutdowns, in that order of priority. Safety takes priority over confidentiality.
Installations are capital intensive, are long-lived. These investments must be protected to guarantee satisfactory return on assets (ROA). This implies providing possibilities for upgrade and modernization for brownfield installations. Implementation of those must be done ‘on-process’. For profitability reasons, CAPEX and project cost should be kept to a minimum.
Other requirements define the capabilities of solutions: optimizing supply chains, reach higher production levels, help personnel improving or complementing skills, or making the use of skills more efficient. (For a more detail discussion of these needs, see our first blog in this series).
Concept
IIoT by Honeywell is a three-tier solution. The first tier is the classical process control with instruments, actuators, a DCS and a control room, with a few additions. First, instruments and control devices, including operator stations are cloud enabled or connected to the cloud via gateways.
Low cost sensors for purposes other than control, for example for monitoring and analysis purposes, can be added and connected via gateways. Collaboration of operations with maintenance, management and engineering on-site, with the help of advanced tools and analytics can further help improving performance. Honeywell proposes access to real-time and historical data with context based on the asset hierarchy, operational performance and business-oriented KPI visualization. Advanced solutions have extensive capabilities to discover abnormal behavior, analyze it, build analytics ranging from simple calculations to complex multi-variable model-based, and create run-time-analytics-based rules automating the early detection of abnormal behavior. For details on the corresponding products Uniformance PHD, Insight, Asset Sentinel and KPI, see “IIoT Analytics Platform Uniformance Suite”. Also operator stations are cloud-enabled, enabling remote operation.
The second tier corresponds to enterprise-wide collaboration. Where cloud-based analysis and expert support was an option in tier one, in tier two it is the only feasible solution. As all advanced solutions are cloud enabled this is technically possible. An important step to make this successful according to HPS is the ‘fusion’ of operational time series, alarm and event information plus other data such as maintenance records, originating from multiple sources in a globally accessible data store. Uniformance PHD can provide this functionality and thereby enable both engineering-science or data-science based analytics at global level.
The third tier extends collaboration of tier two beyond the enterprise boundaries to and ecosystem of third party equipment, engineering, operations or maintenance service providers.
The Cloud platform can be Honeywell’s Sentience platform or a platform of the users’ choice. The platform allows, amongst others, the development of apps by third parties, and which can be made available to users in the tier three scenario. (See also the presentation “Taking Advantage Of New Technologies”, presented at the User Group Conference). An example is HPS’ own connected services to assist users in maintaining performance of systems and applications, dubbed Benefits Guardianship Performa (BGP).
Early Implementations
HPS reports the usage of remote operation of off-shore platforms in the Gulf of Mexico, the remote monitoring of draglines, shovels and trucks in a mining operation in South America, an enterprise-wide implementation in an multi-site refining and petrochemicals company, a predictive maintenance solution in a refinery on the Gulf coast in cooperation with Flowserve (tier three type), an inventory and work-in-progress optimizing involving multiple supply chain partners of a mid-size chemical operation and the optimization of a refinery in the US with help of expert support from UOP.
At the Conference, Ragnar Heksem of Lundin in Norway came to testify about the remote operations center his company installed on shore, to remotely control off-shore installations. Mr. Heksem mentioned the center has been connected using redundant optic fiber and is cyber secure. Up to 180 providers can log in securely. Mr. Heksem mentioned that safety, cost and resource efficiency were the drivers for the project. He said that it is more effective having operators and process experts working physically closely together. As process upsets are often du to human error, the company expects to reduce their frequency. The company aims at having less process upsets because of this. In many other respects remote operation is identical to local operation, Mr Heksem said.
My impressions
From what we know about the solution today, it corresponds to the requirements for Industrial IoT. Built as an extension of existing control systems, it is independent of the function of providing safe and reliable control of critical operations. This protects the existing investments and keeps new investment to the necessary minimum, in particular since cloud solutions are fast and efficient to roll out.
The company has very well described the possible use cases and business processes that end users should take advantage of in designing IIoT implementation.
I believe that the fusion of time series, alarm and event data with unstructured data types enables the user up to take advantage of both engineering and data-science based analytics. HPS provides a very rich set of modeling approaches for analytics that we believe should be capable of creating solutions for virtually any situation, from simple to very complex. The user can choose from several model types, or use them in a cascaded way, complementing one another. The strength of first-principles engineering models is their capability to reliably make predictions in data-scarce regions. These can complement data driven models with faster and reliable convergence for on-line (closed-loop) usage. These two cloud solutions are complemented by real-time models running at the edge that can react on operational signals in split second. The visual analytics approach guiding the user in exploring and building monitoring rules, promises to be a significant source of engineering efficiency. More information can be found here. We believe advanced solutions are likely to enable faster detection, more appropriate decisions, and faster action on early and weak signals, thereby likely contributing to improving reliability and process performance.
Conclusions
HPS created a well-built and articulated IIoT architecture with a rich set of services and apps, built upon Honeywell’s Sentience platform. ARC will continue to follow product development and user applications and report about them.