Industry 4.0 Concepts: Cyber Physical System (CPS), IoT, Digital Twin

The Cyber Physical System (CPS), Internet of Things (IoT) and Digital Twin are all central concepts in Industry 4.0, often used interchangeably in discussions about Industry 4.0 and smart manufacturing. Each refers to a representation of a piece of equipment in cyber space. Such representations are of central importance in Industry 4.0 and for smart manufacturing, since they provide access to real-time operational data of the represented equipment. Use of this data ranges from machine operational status and compiling important KPIs, like OEE, MTBF, MTBA, etc., to big data analytics and machine learning applications, such as predictive maintenance. It is therefore worthwhile to examine what Cyber Physical System (CPS), Internet of Things (IoT) and Digital Twin mean and explore examples of how they relate to each other in manufacturing.

The phrase ‘Cyber Physical Systems’ in manufacturing is said to have been coined for the first time in 2006 by Helen Gill of the National Science Foundation (NSF). The origin of the phrase ‘Internet of Things’ is generally ascribed to Kevin Ashton while at MIT in 1999, whereas the origin of the phrase ‘Digital Twin’ is generally ascribed to Michael Grieves while at University of Michigan in 2001. The phrase ‘Industrial Internet of Things’ was a recent addition to indicate the use of IoT in industrial applications as opposed to consumer applications.

What are cyber physical systems? A CPS in Industry 4.0 is generally defined as a combination of physical (mechanical) components, transducers (sensors and actuators), and information technology (IT) systems (network/communication systems and computation/analysis/control systems). Some cyber physical system definitions in manufacturing include the human, such as the machine operator. In other words, a CPS is a physical world system (machine only or machine plus human) that is connected to the cyber world. A CPS can be either a closed-loop or open-loop system; meaning that it may sense the real-world parameters of the physical system and control it, or it may just sense the real-world parameters and make these available for analytical purposes.

IoT or IIoT is generally defined as a combination of any of the following: trackable objects (such as RFID tags), data objects (such as sensors), interactive objects (such as actuators) and smart objects (such as software components that act on sensor data for any purpose, including pre-processing, control, analytics, etc.).

A Digital Twin is a digital replica of a physical asset. The definition of a Digital Twin emphasizes the connection between the physical and the digital replica, and the data that is generated using sensors. A Digital Twin integrates transducers, artificial intelligence/machine learning, data analytics and context awareness. An example of context awareness is an intelligent thermostat, which senses who is present, so that the person’s preferences for ambient conditions can be taken into consideration.

The CPS concept in Industry 4.0 emerged primarily from a systems engineering and control perspective, whereas the IoT concept emerged primarily from a networking and IT perspective with origins in the RFID context. The Digital Twin concept on the other hand, emerged from an artificial intelligence/machine learning perspective. Nonetheless, all three can be and are being used interchangeably, given that the definitions of the three concepts are converging over time.

GEM Precare agent technology IP straddles all three definitions and therefore GEM refers to CPS, IoT/IIoT and Digital Twin interchangeably. GEM agents acquire data on status, operation, ambient conditions, operator-in-the-loop, as well as other aspects of the operation of a machine, resulting in a multi-dimensional Digital Twin representation of a machine. The agents assign semantical meaning to the data, creating an exact digital replica of the machine’s visible/non-visible signaling interface. GEM agents can be additive to an existing in-the-loop controller, such as a PLC, they can be integrated into the in-the-loop-controller, or they can include the in-the-loop controller.

As an example, consider a wafer pick-and-place machine. The machine picks up wafers from a tray and deposits it on a conveyer belt leading to another machine for further processing. When the tray is empty the pick-and-place machine stops and requests assistance to top up the tray again. The tray being empty is detected via a photo sensor and the pick-and-place mechanism uses vacuum actuation to pick up and release the wafers. A vacuum sensor senses vacuum pressure.

A common situation found is that the photo sensor and vacuum sensor signals are only available to an integrated controller, such as a PLC for instance. A digital representation of the machine would need to include access to these signals. The GEM Precare agent hardware will sense voltage level changes on the photo sensor vacuum sensor signal lines. The GEM Precare software agent will keep count of the vacuum sensor’s signal line changes to count the number of wafers picked up and deposited.

At the same time the agent will monitor the photo sensor’s signal line to signal when the tray is empty. The agent transmits the ‘tray empty’ event with time stamp to the cloud (private, public or hybrid), from where it can be reported in real-time and used to compute MTBA (mean-time-between-assistance). In addition the GEM agent will keep count of the number of wafers that were picked up and deposited, as well as start and end time, by monitoring the vacuum sensor’s signal line voltage level changes. The agent transmits this data to FactoryLogix MES, which provides the contextualization of the data with other data from related processes to deliver more detailed results and KPIs, such as OEE. Installation and getting agents to start streaming data in real-time typically does not require more than 24 hours and doesn’t require the machine to be stopped or powered down.

GEM agent technology IP is application-agnostic and can be deployed in any industrial or consumer application. In addition, GEM has developed market leading subject matter experience in particular in smart manufacturing, with focus on semiconductor and electronics manufacturing.

GEM Precare middleware seamlessly connects GEM Precare agents to the Aegis' FactoryLogix MES platform. This provides manufacturers literally overnight with an upgrade to smart manufacturing without the cost of overhauling their factory floors with new equipment. The combined GEM Precare and Aegis' FactoryLogix solution provides manufacturers with predictive maintenance, MTBF, MTBA, OEE, availability, performance and quality in addition to the extensive MES features available in FactoryLogix.