The IIoT Starts with Sensors

Cliff Ortmeyer, Global Head of Technical Marketing and Solutions Development, Farnell


A key issue is determining the optimum sensor types and how they will communicate their readings to the analytics engines.

Click image to enlarge

Over the last few decades, industrial automation has made it possible for manufacturers and process companies to reduce costs and deliver increased value to their customers. With real-time networks in place, robots, conveyors and process equipment can communicate their status to each other to ensure operations run smoothly.

As cost and time-to-market pressures continue to rise, industrial users need not only greater visibility into these systems; they also need to empower those systems to react intelligently to changes whether due to external forces or internal factors. These are systems in which failure often results in emergency situations, as the ability to detect failures as quickly as possible or even ahead of time is vital. If early warnings of problems can be given, preventive maintenance can be scheduled to prevent downtime and production diverted seamlessly to backup systems. Through real-time measurement of the many parameters that determine the health of the system, the industrial internet of things (IIoT) can transform the enterprise’s ability to react to change.

Sensors provide the insights that power the IIoT and provide the measurements that the IIoT requires. Although many machine tools already incorporate many different types of sensor, the data they provide has often been locked away in silos – used only by the equipment to which they are directly connected.

Real-Time Data

The IIoT opens up the data collected by industrial organisations, bringing together data captured by legacy systems and measurements from dedicated smart-sensor nodes. Through the IIoT, this new generation of sensor node plays a key role in unlocking the power of data an organisation already has but cannot apply in real time. because it lacks context. The data is made accessible by cloud-analytics platforms that can track trends in the incoming data and compare across different data sources in order to compute the best paths of action for a given system.

As an example, consider a fleet of garbage-collection trucks that periodically empty the litter bins of a busy and congested city. Conventionally, the trucks have a fixed daily or weekly route in which each bin is visited and emptied no matter how full it is. In many cases, some bins may be nearly empty while others may be overflowing because the schedule does not take account of how people use them. If an ultrasonic sensor is fitted to each bin, it can at regular intervals send status updates about how much litter is inside.

Software running on servers in the cloud can combine the bin-status data with other information to which they have access such as, traffic congestion in different parts of the city and weather, which can help predict how quickly each bin is likely to fill based on previous readings. However, without the data on an actual bin status, the weather and traffic data are only useful from the perspective of working out which trucks might take longer to complete their routes.

Armed with the live bin-status data, the software can adjust the routes of different trucks dynamically, skipping bins that do not need urgent attention. This can reduce not only the distance covered by a truck but help ensure collection is not held up by traffic. As a result, simply adding one new type of sensor greatly improves responsiveness of the refuse-collection service. Adding location-tracking sensors to the trucks themselves makes it possible to redirect trucks in real time.

In the manufacturing environment, IoT sensor technology can greatly improve responsiveness and flexibility as well as increasing overall uptime through the use and combination of different sensor modalities. In a factory environment, sensors can monitor motor health through analysis of vibration and temperature as well as the fill status of raw materials. Sensors can also guide materials around the shop floor as products are assembled. An RFID label on each pallet identifies the product, so that each tool to which it is directed understands what needs to be done next when its code is read by the tool’s RFID interface.

The Importance of a Cloud-based System

A key issue for organisations that want to enhance their operations through IIoT technologies, is determining the optimum sensor types and how they will communicate their readings to the analytics engines. To maximise real-time responsiveness, not all the analytics will be at the edge. Local gateways will often process the raw data and relay instructions to nearby actuators and other equipment to keep the system running smoothly. In parallel, they will filter and organise data so that it can be ingested by remote cloud analytics engines more easily.  The cloud makes it easy for organisations to access high-performance computing on demand to deliver real-time knowledge and maximise the value of the incoming data.

An example of a cloud-based system for handling IIoT data is Schneider Electric’s EcoStruxure. The solution consists of apps, analytics and services layers that have been tuned for industries; including building management; data-centre operations; industrial control; and energy-grid networks. As interoperability is key to supporting the diverse hardware and systems that are used in each of these four industries, EcoStruxure enables a breadth of device-agnostic applications, analytics and services for seamless enterprise integration.

Click image to enlarge

Figure 1: EcoStruxure

EcoStruxure includes support not just for processing in the cloud but also in the edge-control layer, providing the critical ability to manage operations locally and support advanced automation and operator capabilities. Local servers will act as wireless gateways for the sensors that are closest to them. However, with support for low-power wide-area network (LPWAN) technologies such as Sigfox and LoRa, a single gateway can cover a very large area, enabling IIoT support for users in energy-grid and utility-supply applications as well as agriculture. Security monitors on lock gates on remote canals can be as accessible, as the door-status monitors at a factory or substation can extend over tens of kilometres thanks to Sigfox’s and LoRa’s RF communications range.

Click image to enlarge

Figure 2: Grid

Inside a campus or building the Schneider Server platform provides support for sensors that use Zigbee and similar wireless local-area network (WLAN) technologies. To provide maximum flexibility, the XIOT sensors from Telemecanique can be configured with matching transceivers to support any of these WLAN or LPWAN protocols. The use of low-power networking protocols makes it possible for the battery-enabled sensor units to perform measurements and transmit data they capture. They can perform these tasks over periods of five to 10 years with no need for external power onsite. Each sensor contains its own processing logic so that it does not transmit more data than is necessary to keep the servers up to date. Threshold conditions and other alarm types, such as presence or absence of a key setting, can be programmed into the sensor. Numerous types of wireless sensor can be used on the network. They include safety switches to monitor potentially dangerous equipment, RFID interface modules, ultrasonic sensors, pressure sensors and position monitors.

Click image to enlarge

Figure 3: Building

Many industries can take advantage of the combination of XIOT sensors and the Ecostruxure cloud. In one application, sensors were installed in a utility’s ducts and drainage pipes to monitor whether valves used to help convey rainwater through the network are functioning as expected. No electrical connection was available although photovoltaic panels could be used in some locations to power the sensor units. As well as providing analytics to users, the installed system based on 1000 sensors was able to save more than €100,000 per year through the avoidance of flooding and reduced manual checks on the valves.

In agriculture, limit switches fitted to mobile irrigation systems are now used to detect obstacles in the path of each machine and to know when each irrigation cycle is finished. Pressure sensors check for broken or leaking pipes and the data is relayed over LPWAN, so that a single cloud system can monitor the progress of irrigators operating over a huge area without requiring costly regular visits by human operators.

The approach applies just as well to resource extraction, with limit switches and other sensors used to monitor the lengthy conveyors used to carry rock ore around an open-cast mining operation. Key parts of the mining site can be secured with door-position and RFID sensors. If someone enters an area without the right RFID authentication or at an unexpected time, analytics software can detect the anomaly and raise an alarm. This results in a much better insight to activity on sites that, previously, have been extremely difficult to monitor. Putting these different sensors together provides industrial users with a much more comprehensive and timely view of their operation, unlocking the power of the IIoT.