Embedded Files
ISA95 standard
Relation of the communication between the different levels of ISA95:
layer 4: Enterprise Resource Planning (ERP)
layer 3: Manufacturing Execution System (MES)
layer 2: Supervisory Control and Data Acquisition system (SCADA)
layer 1: industrial controllers
layer 0: mech
Examples of use: power generation, water treatment and distribution, wastewater collection and treatment, managing oil and gas pipelines, electric power transmission and distribution, and managing wind farms.
Layers 0 and 1 are operational level layers and the layers 2, 3 and 4 are management level layers.
The vertical communication between a components in layer 0 and the belonging controller in layer 1 is hard real-time. The vertical communiction in the upper layers is not real-time. The horizontal communiction between different machines in layer 0 and different controllers (machine to machine) in layer 1, can be real-time and not real-time.
Protocols for the horizontal communication in layer 0 and layer 1 real-time:
– Ethercat Automation Protocol (EAP)
– UDP
and not real-time:
– OPC UA
– ADS
– TCP/IP
The sample frequencies could be some kilohertz. So a fieldbus is necessary:
– Ethercat
– Profinet
– CAN-bus
– Ethernet/IP
Enterprise resource planning (ERP) is an often real time computerized system for integrated management of main business processes.
Manufacturing Execution System (MES) is a computerized system used in manufacturing to track and document the transformation of raw materials to finished goods. The MES is responsible for planning and co-ordination of the production order and for the production logging (e.g. start time, number produced, production time). MES works in realtime to enable the control of multiple elements of the production process, like: input, machines, service support and manpower. MES is an intermediate step between the enterprise resource planning (ERP) system, and a supervisory control and data acquisition (SCADA) or the process control system. MES sent to PLCs: work instructions, recipes and set points and receive from PLCs: process values, alarms, adjusted set points and production results.
Supervisory Control and Data Acquisition (SCADA) is a control system architecture for high-level process supervisory management. SCADA comprise computers, graphical user interfaces (GUI), networked data communications and peripheral devices like: programmable logic controllers (PLC) and discrete proportional-integral-derivative (PID) controllers. An important part of most SCADA implementations is alarm handling.
An edge device at SCADA level could be used for remote connection for users from other companies. Those companies don’t get directly access to the industrial controllers and the mechanics. Then at SCADA level we need to implement:
– Internet connection
– Internet router
– firewall
– VPN
– Transport Layer Security (TLS)
Smart Industry technology
With technologies as IoT, cloud, machine learning and artificial intelligence we can make (business) processes smarter. More efficient, flexible, sustainable, a better customer satisfaction and financial performance. The data can be collect automatically with sensors and can be used to identify the imperfections. With machine learning and artificial intelligence we can then improve the imperfections (process optimization). But of course the slowest production process determine the speed. All the processes needs to tune up with the slowest process to avoid large work-in-progress buffers or a crash.
Local cloud
For economical energy control it is important to keep the machines and agv so simple as possible by moving the heavy things to a edge computer or to a cloud. Also in a lot of situations you need information from other machines. In the cloud all the information from all the machines comes together. A cloud is e.g. usefull for:
- data streaming (e.g. machine data or video streaming)
- machine learning-algoritms for solving bottenecks and other problems.
- artificial intelligence (AI)
- augmented reality (AR) and virtual reality (VR) for support at distance
An agv (automated guided vehicle) is a logistical robot (e.g. wheel based or carterpillar based inspecting robot, inspecting drone) powered by fleet management software in the cloud. E.g. an inspecting robot or inspecting drone with a heath camera to discover gas leaks.
To get not to much cables mostly a wireless connection with a low latency is used. A 5G network has a latency roundtrip of 1-10 ms. So a network with 5G Small Cell indoor antennes is a good sollution.
To protect the production data it is usefull to use a local cloud (own local frequenty without using a telecom operator). E.g. they Ixon Cloud 2-platform with:
- well-ordered and intuitive interface
- there own iot-portal
- there own dashboard
- Ixon Studio for creating and adapting they portal and dashboard
- connecting with the machines
- data collection
- allerts
- solving problems at distance
- full control of user access en user rights
- role based user control
When the speed and security is realy essential then a cloud is maybe not a good solution. Then maybe edge AI can be usefull. It’s faster and more secure but it has some problems with energy control and calcultion:
– storage signals of all sensors
– synchronisation of sensor data
– compression of sensor data
– combination of sensor data
– extract to a format for steering and decision algorithms
Also at local level ( edge AI) the information from other machines is not available (a leak of information).
Machine learning
The sensor data from machines and robots is storage in the cloud. This data can be used by algorithms for interpretation and understanding.
The cloud and deep learning can be used for all sensors that needs compression and information extraction, e.g. radar, lidar and sonar.
Two ways of improving the machines and robots with an algorithm (AI):
– learning by examples
– reinforcement learning
With AI the rules are chanching:
rules + data => classical programming => answers
data + anwsers => machine learning => rules
PLC
PLC or Programmable Logic Controller is an industrial microprocessor-based controller (computer) that has been ruggedized and adapted for the control of manufacturing processes. It has programmable memory used to store program instructions and various functions. A PLC works in a program scan cycle. It executes its program repeatedly (read inputs, execute the program, write outputs).
A PLC consists of:
- a power supply unit (converts AC voltage to DC)
- a CPU (execute the program, interprets inputs and sends output signals)
- a memory unit (program, input data)
- an input interface (where the controller receives data)
- an output interface (where the controller sends data to external devices)
- a communications interface (receive and transmit data on communication networks)
Traditional PLC's are generally closed platforms. Some PLC with an open platform:
- ctrlX from Bosch Rexroth
- PFC200 from Wago
- PLCnext from Phoenix Contact
- Revolution Pi from Kunbus
Types of PLC's:
- traditional PLC (CPU, internal memmory, storage e.g. MMC-card and own cover)
- Slot-PLC e.g. PC PCI-card
- soft PLC e.g. software for PC or embedded PC
- OPLC or operating panel PLC, PLC with HMI (keyboard, display or touchscreen)
- safety PLCs, suitable for safety-critical applications (SIL standard)
Interfacecard or moduls
- digital card/module
- analog card/module
- precision timing module/counter module
- communication module (e.g. RS232, RS485 or ethernet)
Fieldbus protocols
- Modbus-protocol
- Modbus Plus
- Profibus DP
- Profibus PA
- Foundation Fieldbus
- AS-Interface (Asi)
- CAN (Controller Area Network )
- Canopen
- HART (protocol)
- Interbus
- Modbus/TCP
- Profinet
- Ethercat
- IO-Link
- Sercos-III
- Varan
- Ethernet/IP
- Powerlink
Google Sites
Report abuse