Level-5 automation architecture is a very well-known terminology in the world of automation.  With ISA-95, it is found that most facilities of process industries have achieved almost complete automated production process, where assets are connected to a central control system (i.e. Digitally distributed control systems [DCSs], Historians/manufacturing execution systems (MESs), and supervisory control and data acquisition [SCADA]).

However, ISA 95 defines the process industry automation hierarchy. The ISA standard establishes a model for the exchange of information between the system in five levels:

1. Level 0 – Field Instruments
2. Level 1 – PLC Control systems (PLC)
3. Level 2 – Process Control System (SCADA)
4. Level 3 – Manufacturing Execution System (MES)
5. Level 4 – Enterprise Management Portal (ERP)

Hence, the need of the day is the integration of Design Engineering Data with Process Automation.

Industry 4.0 and Design Engineering Industry

This article covers Industry 4.0 integration with Design Engineering Data enabling plant automation and Industry 4.0 influence on existing Process Automation. We have seen many process industry personnel struggling (especially in the case of large oil refineries, chemical and steel plants) to order new instruments valves, drive mechanical equipment cables, and all other necessary accessories because of poor preservation of design engineering documents developed during the design phase of the plant.

Before the 1980’s design engineering was mainly dependent on manual drafting and Typewriter. Development of engineering drawings required engineering tools like T-square, drafting board, French Curve, Rulers, Compass, etc. It took a lot of time to develop an engineering drawing and deliver.

In 1989 Autodesk first released a desktop version of AutoCAD (Commercial computer-aided design) software, which led to revolutionary changes in the design industry and expedited the creation of engineering documents. From 1990 onwards, the design engineering time was reduced exponentially by using software tools. At that time, AutoCAD was used primarily by engineering disciplines like Civil, Mechanical, and Electrical.

However, Instrumentation did not have any specific software tool for developing Piping and Instrument Diagram, Instrument Index, Signal List, Instrument and Valve Data Sheet and Logic Diagram and Cable engineering, etc.

The first popular instrumentation software, ‘In tools’, was initially developed in Israel and was bought by Intergraph and renamed ‘SPI’. It is an engineering tool that uses ODBC compliant databases such as Oracle, MS-SQL, and Sybase. It is being integrated with other engineering tools to create a data-centric engineering environment.

Another popular plant engineering software is COMOS (Component Object Server). It is an open architecture that allows different engineering data formats from various data sources to integrate effectively.

Industry 4.0 standards depict an integrated information platform where all the engineering documents can be linked and accessed. The plant operator can now access process flow diagrams, piping engineering drawing data sheets of all mechanical, electrical, and instrument components, and any equipment/instrument ordering information by simply clicking the particular display available in the plant HMI.

Intelligent Software lifecycle allows all plant life cycle phases to have comprehensive information management, regardless of functional assignment. This shall get maximum production profit from maximum reliability in decision-making and worldwide access to documents and data.

Fig (a). COMOS screenshot of 3D walk view of the process plant and related engineering drawing developed during the design phase

Industry 4.0 and Process Engineering Industry

Today, the Government’s Technology Strategy focuses on using information and communication tools and improving production/manufacturing facilities. For Industry 4.0 Implementation, the Internet of Things is the key enabling technology.

The final goal is to achieve an intelligent factory in all sectors that evolves with resource efficiency, adaptability, and ergonomics, including the integration of technology partners and customers in both business and value processes.

IoT relates to manufacturing through automation and management of machines and systems using the Internet. Connecting processes, monitoring, control, and management of these procedures allows manufacturers to manage and access their factory operations better by allowing different systems to ‘sync with each other and understand’ the needs of each system and ensure processes keep running smoothly.

Using data collection and processing in the application context, the systematic use of best practices to improve profitability and safety independent of expertise and people.

A connected network of sensor-built monitoring and production software and equipment provide actionable information and enable informed decisions. Access to this knowledge is not restricted or bound by a human connection to the facility where the actual equipment operates.

IIoT is a network of intelligent components like hardware, software, and services that shall facilitate democratising input data from the edge of the network at the field level or distant point for analysis, performance and to attain a successful collaboration. The key to this effort is effectively utilising all prevalent enabling technologies, both information technology (i.e. software, network, protocols) and operation technology (industrial validated, reliable, and digital interfaces with the field devices) to get the desired result by working together.

IoT incorporates big data and machine learning, harnessing the machine-to-machine communication, sensor data, and automation technologies leading to sustainable practices, improved quality control, supply chain traceability, and overall supply chain efficiency in industrial manufacturing.

Predictive Maintenance for Industry 4.0

Predictive maintenance is a method to prevent asset failure by analysing events before they happen by analysing online process parameters/production input data and identify recurring/familiar patterns and predict the problems/issues. Till now, maintenance personal carried out maintenance work as per schedule and repair equipment regularly, and prevented downtime/failure leading to unnecessary use of resources and productivity losses. Production /Process parameters of assets are obtained from the field devices connected to a centralised data server/cloud by utilising validated/proven communication protocols with gateways.

Data about management areas are obtained from MES and ERP systems, along with production process flows. They later shall be integrated into the centralised data server/cloud to provide context to the process parameters asset data. At the last stage, predictive analytics algorithms are applied, and these offer necessary solutions to reduce downtime, which is generally diagnosed using software for root cause analysis.

Two common approaches used in predictive maintenance are Rule-based and Machine learning-based. Ruled-based predictive maintenance is commonly referred to as Condition Monitoring systems. The crucial input to rule-based predictive maintenance is from the field sensors to continuously collect data on assets. It sends alerts as per programmed logic/rules, including when a predefined setpoint limit is achieved. In rule-based analytics, the design group shall work alongside detailed engineering and service groups to prove the actual causes or contributing factors leading to the equipment’s failure.

After common reasons for equipment, failures are identified and proved, equipment manufacturers shall build a virtual or digital twin model of their connected system. Later the related product is defined using cases with “if-this-then-that” rules wherein they describe the inter-dependencies between the IoT systems and their behaviours. For example, suppose the shaft speed and its temperature have reached a specific predefined level limit. In that case, the control system which is preconfigured or programmed shall send a caution/alert message to an operator console to review and attend to the problem ahead of failure.

Machine learning predictive maintenance looks at a massive volume of test data and history values combine them with structured machine-learning (ML) algorithms to run and get various options/scenarios. Later they also predict suitable options to show that what went wrong in equipment and when. One real-life example of machine learning would be social media. When we upload a photo on social media, it recognises a person and suggests a familiar friend. To make these predictions, it uses information from databases like friend-list, images available, etc., and it makes predictions based on that.

With the help of machine learning, systems make accurate decisions at high speed. This technique can analyse large and complex data and are inexpensive.

Cybersecurity Considerations

The main aim of any new technology is to enhance productivity and increase the efficiency of the process cycle of the plant and ensure no stoppage. Unfortunately, all these cannot be realised without of implication of cost and cybersecurity risk. Hence, for these above reasons, Ethernet-based devices must include safeguards in their products to ensure that network bandwidth is safe so that viruses or malware cannot be loaded to the device. In the same way, unauthorised reconfiguration of devices cannot be done, unwanted access is not given, unauthorised reconfiguration of devices will not be permitted, and unauthorised writes to memory spots are not allowed by the device. In addition, the physical security of these devices requires that they are accessed by authorised personnel only. The process data in the machine, if not used for control, shall have only read access. It is essential that the entire product life cycle, including design, build, and test adheres to tight process and quality assurance requirements.

The main challenge for implementing cloud-based architecture is to ensure the hundred-person security of process data. At a minimum, a two-layer protection scheme includes software and hardware restricted access and should be installed in the device. Additionally, post-installation support should be considered to assist on-site protection of site data and property.

Conclusion

Automation Industry 4.0 with a digital outlook offers the next step in productivity and efficiency by exploiting today’s available advanced technologies, like  IIoT, 3D printing, artificial intelligence, robotics, wireless,  and others that will enable the industry to achieve the best targeted, optimised levels of productivity and efficiency.  Finally, combining all these new evolving technologies will provide a technological improvement/innovation/development that will represent or provide a breakdown of benchmarking and the creation of new startup/business ventures and lead to changes in the relationship between producers and customers in the future.

References

1. Industrial Field Instrument Progression To Smart Digitization Enabling IIOT by Ms Latha D S, Tata Consulting Engineers Limited.
2. Bridging the gap between HART devices and IIoT by Tina S. Todd
3. https://www.seebo.com/predictive-maintenance/- Industry 4.0 Predictive Maintenance
4. https://www.isa.org/intech/201806web/ – How do you define IoT and Industry 4.0 as it relates to industrial manufacturing? by Bill Lydon