What is a 'Smart Connected Sensor'?

In addition to the actual measured variable acquisition, such an SCS contains the complete signal preparation and signal processing in the same housing. It usually has a digital interface (e.g. Modbus, CAN, CANopen, IO-Link, Ethernet) for communication with higher-level systems. The market even offers variants with an integrated mobile radio modem.

However, an SCS always includes a special cloud service platform to which the sensor can transfer data without the need for additional engineering. A virtual data image of the sensor is created on the service platform and supplied with current data. In addition, an API is available for authorized access to the data image. Valuable additional functions can be implemented via the cloud, such as comparing the measured variable received from the sensor with an IT database in order to place the measured variable in an application context and send an alarm or notification if required.

An example of this could be a fill level sensor in a system that sends the measured fill level to a cloud service platform on the internet whenever there is a change. There, the measured value is received by a software component assigned to the SCS and checked against certain limit values. If, for example, the value falls below the minimum fill level, the service platform sends a refill notification to the ERP system of a service partner to initiate a container refill.

Direct or indirect cloud connection

An SCS can communicate with the cloud service platform in different ways. In the simplest case, the sensor has an integrated 2G/3G/4G cellular modem with SIM card and can reach the cloud via the cellular network of a network provider. This solution enables complete pre-configuration ex works so that the sensor only needs to be installed in the field. No further on-site configuration is required, so mass roll-outs would be possible without any problems. An integrated Wi-Fi interface is also conceivable.

Images

An SCS with a direct Internet connection delivers measurement data to the cloud service platform via TCP/IP or using a Short-Range Wireless Network (SRWN).

© SSV

In this case, however, the SCS must at least be configured locally for the respective 'Wi-Fi access point', which also requires a special configuration interface. In both cases (mobile radio and Wi-Fi), a complete TCP/IP stack and special security modules are required directly in the SCS to defend against cyber attacks. However, a 'wireless sensing' SCS variant is also possible, which communicates via a short-range wireless network (SRWN, e.g. ZigBee) with a special gateway that forwards the sensor measurements to the cloud service platform. In this case, TCP/IP plus security is only required in the gateway. The individual wireless sensing node would then be much cheaper to implement.

For many applications, it is sufficient if the sensor only has an inexpensive Bluetooth Low Energy (BLE) interface and is supplied together with a smartphone app. The sensor itself then has no direct connection to the cloud. This is realized via the app. The app can pre-process, modify and temporarily store sensor data and visualize it on site. For example, it is possible to visualize the actual status by directly querying the sensor measured variables using BLE. At the same time, a history can be displayed by requesting historical data for the relevant sensor via the cloud service platform.

SCS as a data supplier

Valuable data can also be obtained in existing machines and systems by retrofitting suitable SCS. Forwarding the sensor data to a cloud service can enable condition monitoring, for example.

© SSV

A business process can be automated with a single SCS - for example as a fill level sensor. Several SCSs installed in a machine or environment even enable service-oriented business models. However, this requires extensive data images and long-term data storage (history).

If, for example, the failure of individual system components is to be predicted for a predictive maintenance service with the help of SCS, a corresponding SCS is required at all neuralgic system points. All sensors together then provide a condition monitoring image in the cloud. In order to apply the predictive analysis tools of a cloud service platform to the SCS data, the most comprehensive and meaningful condition monitoring history possible is first required. Machine learning can then be used to derive a prediction model from this data, which can predict the probability of failure of individual assemblies using predictive analytics. Even changes in the quality of the items produced can be predicted if, for example, the tools used are included in the SCS data image.

Developing digital business models

Suppliers of sensors, machines or systems should immediately develop a digital business model and start offering SCS extensions with data-based services for existing products as soon as possible. Service version '1.0' does not necessarily have to be perfect. The winner will be the one who can generate valuable information from data and recognize further patterns for their own customers. This will certainly require several iterations.

Author: Klaus-Dieter Walter is a member of the management team at SSV Software.

Intensive seminar on the topic of IoT

The seminar with Klaus-Dieter Walter from June 8 to 9, 2016 in Munich provides comprehensive answers and examples of both consumer electronics-oriented IoT and the Industrial Internet of Things (IIoT).

The combination of theory and practical exercises, which will be carried out using an IoT developer kit, will give participants a thematically structured insight into the subject matter. The examples range from smart homes to smart factories. Further information can be found at: www.training-for-professionals.de.