Manufacturing processes are being digitalized and automated using Automation Systems. But Automation Systems are complex and they contain lots of under-used equipment. The cost and time of manufacturing can be reduced if the under-used equipment on Automation Systems is used e ciently. Discovering the existing functionalities on an AS and installing new functionalities with low e ort plays an important role to e ciently use an Automation System equipment. In this demonstration, we will present an approach employing Web of Things and Semantic Web technologies, to make an Automation System equipment transparent, and present how the e ort of installing new functionalities on the eld devices can be lowered if the Automation System equipment is transparent.
Industry 4.0, which is also referred to as the fourth Industrial revolution aims at
lowering the time and cost of manufacturing individualized products [1].1 But
an Automation System (AS) used in a manufacturing process posses under-used
equipment, which when used e ciently can reduce the cost and time of
manufacturing products. This poses certain challenges as Automation Systems (ASs)
are complex, they contain lots of capabilities, con gurations and functionalities
running on them, which should be taken into consideration while installing new
functionalities on them. Therefore, it is a challenge to: (
We developed the architecture as shown in Figure 1a for semantic-based engineering of ASs. The rst component is a semantic repository with RESTful interface, which is used to store semantic descriptions of new functionalities. The second component is semantic-based engineering tool that, provides a graphical user interface to do engineering with our semantic-based approach. The tool also provides an easy access to the repository and an AS to discover its equipment and functionalities or to deploy new functionalities on an AS.
The third component is the device architecture. The device here refers to an edge device that is embedded into an AS, which controls the operation of eld devices (sensors and actuators) of an AS. Each edge device is endowed with a Micro-reasoner that empowers it to interpret TDs of the eld devices and semantic models coming from the engineering tool, process the data from the eld devices or to control them locally on the AS itself. Micro-reasoner as shown in Figure 1b is o ered as a RESTful Web service. It consists of two components: Micro Event Engine implemented in C and a the Datalog reasoner to do reasoning in datalog [7], which is an open source C and LUA4 implementation.5 Micro Event Engine, which is based on the work from Anicic et al. [6] uses event rules to do Complex Event Processing (CEP) of eld devices. 3 http://mqtt.org/documentation 4 https://www.lua.org/ 5 http://www.ccs.neu.edu/home/ramsdell/tools/datalog/datalog.html
In this section we describe a use case for re-engineering an AS on the FESTO6 Process Automation workstation. The workstation consists of two tanks, there are various sensors and actuators attached to the tanks to monitor their state. Among others, the workstation consists of an ultrasound sensor that measures level of liquid in a tank, oat sensors, a pump and a pneumatic valve. We attached a micro-controller to each sensor and actuator of the workstation where the TD of the eld device is stored and accessed. In this demo, FESTO workstation is equipped with three SIMATIC IOT20407, which are used as edge devices.
In our use case, initially eld devices on the workstation are engineered to ensure over ow protection on the upper tank, using the oat sensor on the upper tank and the pneumatic valve, such that when the oat sensor detects over ow of liquid in the tank then the pneumatic valve lets liquid from upper tank to the lower tank. In this settings, if the oat sensor is malfunctioning then, over ow protection cannot be ensured on the upper tank until the malfunctioning oat sensor is replaced with a new sensor. In this demonstration, we will present how the ultrasound sensor deployed on top of the upper tank can be re-engineered with less e ort, to be used in place of the oat sensor to ensure over ow protection on the upper tank. 4
A TD is created for each edge device, which describes the properties, events and actions of eld devices controlled by the edge device. Figure 2a shows a snippet of a semantically enriched TD, which controls the ultrasound sensor. The TD is then converted to datalog facts and stored in the Datalog reasoner on the IOT2040 as shown in Figure 2b. In a similar fashion, new functionalities and their requirements are described semantically and stored in the repository. An example of a functionality used in this use case is: a eld device should detect over ow status of a tank, if the level of liquid in a tank is over certain threshold. The requirement to install such a functionality on a device is that the device should be capable to measure level of liquid in a tank.
Having semantic descriptions of devices and functionalities, then re-engineering
is done using engineering tool in the following steps: (
context.jsonld", "http://SWAS/interactions/interactions
context.jsonld"], "qu":
"http://purl.oclc.org/NET/ssnx/qu/qu#", "ssn":http://purl.oclc.org/NET/ssnx/ssn#, "schema": "http://schema.org/", "eclass": http://www.ebusiness
unibw.org/ontologies/eclass/5.1.4/#, "name": "MyUltrasonicSensor", "@type": ["ssn:Sensor",
"eclass:C_AKE655002-tax"], "uris" : ["coap://192.168.2.82:5683/ultrasonicSe nsor", "http://192.168.2.82:8080/ultrasonicSen sor"], "encodings":["JSON"], "ssn:onPlatform":"Tank1", "ureasoner" : "true", "properties":[ { "@id" : "level", "@type": http://SWAS/interactions/liquidLevelProperty, "name" : "liquidLevel", "valueType" : {"type" : "float"}, "writable" : "false", "qu:unit" : {"@type" : "qu:millimetre"}, "schema:minValue" : "0.2", "schema:maxValue" : "800", "hrefs" : ["liquidLevel"] } ]} b. Ultrasonic Sensor Datalog facts name("td","MyUltrasonicSensor"). ureasoner("td","true"). uris("td","coap://192.168.2.82:5683/ultrasonicSensor"). onPlatform("td","Tank1"). properties("td","propetery1") name("propetry1","liquidLevel"). hrefs("propetry1","liquidLevel"). @type("property1",http://SWAS/interactions/liqui
dLevelPropetry"). functionality and converts it into an event rule executable by Micro Event Engine. Therefore the functionality of malfunctioning oat sensor is replaced with ultrasound sensor using semantic-based approach.