Gaining the upper edge in the Industrial Internet of Things

The Germans have coined the term ‘Industry 4.0’ for computerisation of the workplace; this is also frequently referred to as the ‘fourth industrial revolution’. Collaborative robots are helping drive the move forward – but so too is the Industrial Internet of Things (IIoT). 

It’s hard to escape the reach of the Internet of Things (IoT); along with the cloud it seems all pervasive. But where the IoT’s cloud can appear to be somewhat fluffy, the IIoT’s relationship with the cloud is better defined. While the IoT is fully immersed in the cloud, for the IIoT, most of the action is at the edge.

Processes in the IIoT are time critical. Decisions need to be made in real time to keep systems running and continually refined. For this reason, most data collected at the edge is processed there. Connections need to be low latency to avoid any processing delays occurring, and data integrity from edge nodes is critical. Connectivity requirements can differ, based on a number of parameters such as range, power, security and reliability.

For the high reliability typically demanded of IIoT systems, Ethernet offers the most robust (wired) connections, far more so than the wireless networks that are typically used to transmit IoT data. There are various other choices available, such as the slower speed twisted copper pair KNX network. There is generally a gateway device that controls other edge devices and also relays information to the cloud. Examples of protocols used for satisfying the requirements of industrial control and automation applications are PROFINET and EtherCAT. Ethernet speeds can be up to 100Gb/s.

The wireless networks into which the IIoT edge nodes or gateways also need to connect are available with a wide range of protocols. Choice of protocol is dependent on various requirements, notably signal range. Other considerations are the topology of the network itself – how it is configured – and the overall power budget of the application. Protocols vary from those for short range – such as Bluetooth or radio frequency identification (RFID) for personal networks, to those for mid range applications, for example, WiFi for wider area requirements, right up to cellular based wide area network (WAN) protocols such as narrowband (NB)-IoT and the low-bandwidth second-gen LTE-M.

IIoT nodes are required to sense and actuate as well as communicate. While everything may be digital on the cloud side, the side of the gateway that fronts onto the real world is likely to be faced with a mixture of different componentry of mixed heritage and, indeed, age. There may well be a combination of components for the gateway to deal with, including both analog and mixed signal features, sometime housed on the same chip. 

A heavyweight gateway with requirements for processing in real time may rely on a specialised operating system such as Linux that has been designed for use by embedded systems. Other devices on the edge may use a less resource heavy real time operating system such as Keil RTX from ARM or FreeRTOS. As well as functionality requirements, the choice of operating system will depend on the availability of resources in the device. The hardware used is also often constrained by features such as cost, space and power budget. 

This may seem like a bewildering choice of alternatives, but there are a lot of widely differing requirements out there. There are also off-the-shelf modules available that offer mixed communication potential as well as mixed signal capability to provide interfaces for actuators and sensors, but this can result in a bulky solution that does not actually represent best fit for the requirements and is not as effective as might be desired. 

Where there is a need for a combination of mixed signal and analogue components, it may be better to opt for a system-on-a-chip (SoC) design such as those available from S3 Semiconductors. The company can integrate an analogue front end (AFE) with calibration and compensation algorithms, control functionality, communications protocols, security functionality and embedded software onto a single chip to provide precisely the requirements you need. 

Along the way, customers can save themselves up to 80% in terms of their bill of materials (BOM) costs and 75% in terms of PCB area and power. That’s definitely going to provide them with the edge they need.