Early-stage IoT concepts defined sensors that linked directly to the cloud. However, as vertical industries started seriously evaluating IoT architectures to extract greater business value, it became clear that this one-size-fits-all approach was impractical for various reasons.
Consider just a few of the implications of a cloud-first model in Industrial IoT (IIoT) deployments:
• Data and device security: The potential of insecure endpoints communicating directly with the cloud meant hackers could exploit vulnerabilities to access sensitive industrial networks.
• Runaway networking costs: Sensor-to-server data transmissions (especially over public networks) can be so costly they prohibit scaling to the thousands of nodes required by many IIoT deployments. Add large volumes of measurement and status data generated by industrial sensors, and network congestion, packet delays, and inefficient bandwidth usage abound.
• Power consumption of always-on sensor nodes: Remote sensor nodes require continuous connection to the network and an energy source. This is particularly challenging in remote settings like mining and agriculture, where limited access can mean replacing batteries or troubleshooting networks, costs thousands of dollars.
New classes of secure hardware, networking, and battery technology emerged from these challenges to redefine how IoT systems were architected, and industrial devices were designed. The technology revolution began by combining security and energy efficiency in edge-centric silicon.
The low-power foundations of IoT processors
Introduced as real-world IoT requirements were being defined in 2009, Arm Cortex-M0 CPUs offered the ability to operate solely on 16-bit ‘thumb’ instructions rather than the 32-bit instructions required by its predecessors.
Thumb instructions’ compact encoding method enables code density improvements of roughly 30% on processors like the Cortex-M0, which has a cascading effect on memory usage (lower), die sizes (smaller), power consumption (less), and ultimately cost (reduced).
Fast-forward to today and devices based on the Arm Cortex-M33 architecture feature thumb instructions and built-in hardware security via features like TrustZone.
TrustZone delivers hardware-based data and device security through a secure root of trust (RoT). When combined with the energy efficiency of Cortex-M33 CPU cores, TrustZone creates secure, battery-powered IoT devices that can operate for extended periods in remote settings. It also doesn’t detract from CPU performance, as Cortex-M33 processors deliver an impressive 1,5 DMIPS/MHz and 4,09 CoreMark/MHz for handling complex tasks at the edge to reduce reliance on centralised cloud processing.
From the beginning of IoT rollouts through to today, Cortex-M-class chips continue to deliver possibilities for various IoT use cases.
The rise of LPWAN
The success of energy-efficient IIoT edge nodes is not only a result of their host processor but also how they connect. In the late 2000s, the advent of 4G technology signalled the decline of earlier networks, highlighting the need for a new low-power, wide-area networking (LPWAN) technology that facilitates long-range communication for IoT devices.
LPWAN technologies such as LoRa have emerged as an appealing method for linking battery-powered IoT devices to networks. Its long-range capabilities and low energy consumption make it an ideal choice for IIoT applications like asset tracking, environmental monitoring, industrial automation, smart agriculture, and smart cities.
Today’s LoRa transceiver modules facilitate LPWAN communications over distances of up to 15 km, while consuming approximately 40 mA of current during transmission. Typically, LoRa modules interface with host processors like Cortex-M-class devices through UART, and communicate via ASCII commands, streamlining integration with IoT devices.
These transceivers pair with sub-GHz antennas that meet the frequency requirements of LPWAN networks, many of which are available in compact SMD form factors that fit the space constraints of edge devices. In addition to supporting protocols like LoRaWAN, some of these antennas also support short-range wireless technologies like Wi-Fi, Zigbee, and Bluetooth to enable the creation of backhaul-enabled wireless sensor networks.
Lithium battery technology advances for IoT edge nodes
Thanks to the availability of secure, energy-efficient computing technology and LPWAN networking, the idea of battery-powered IIoT sensor nodes became a reality. The IIoT industry embraced the concept of battery-powered sensors, and demand for dependable, high-density power sources increased.
Lithium-ion batteries emerged as the preferred choice for powering these sensors, thanks to consistent power density and reliability improvements. These advancements yielded the ability for IoT devices to operate for extended periods on a single battery charge – a critical requirement for many agriculture, mining, and industrial applications. Meanwhile, the improved reliability of lithium-ion battery technology led to reductions in maintenance and operational expenses, while ensuring uninterrupted data collection and communication.
A Qoitech study on the compatibility of LoRaWAN technology and coin cell batteries highlighted the pairing’s potential in enduring, low-power wireless IoT sensor nodes [www.qoitech.com/blog/lorawan-iot-coin-cell-battery]. In the study, researchers tested the performance of coin cell batteries using a battery-profiling tool. The tool measured a 40 mA
This harmony between LPWAN technology and high-density lithium-ion batteries has helped propel the IIoT landscape, enabling new energy-efficient wireless sensor nodes. Lithium coin cell batteries have emerged as the go-to power source for these devices due to their compact size, impressive energy density, and extended lifespan. The availability of diverse lithium coin cell battery options – available in various chemistries and configurations tailored to specific IoT applications – gives developers freedom of choice.
Future of technology for the Industrial IoT
Recent IIoT technology advancements have not been limited to the edge; they’ve also extended to the control layer. These improvements have led to multicore systems-on-chips (SoCs) featuring multiple CPU or graphics cores, integrated neural network accelerators, and dedicated IP blocks for executing analogue, security, and other workloads.
These high-performance chipsets almost always contain multiple high-speed I/O interfaces that streamline system integration in a number of deployment contexts. They are also candidates for embedded virtualisation using technologies like hypervisors and single-root I/O virtualisation (SR-IOV) that partition on-chip cores, memory, and I/O resources. As a result, multiple mixed-criticality workloads can run and execute simultaneously on a single physical processor, maximising resource utilisation, and reducing overall size, weight, power consumption, and cost versus multiprocessor solutions.
Elsewhere, networking standards like Ethernet Time-Sensitive Networking (TSN) are rising. TSN introduces deterministic communication capabilities from the control layer to sensor nodes and enterprise systems for fine-grained timing control, precision device management, and task-oriented workflows like virtual programmable logic controllers (vPLCs). The convergence of these technologies is expanding functionality as IIoT nodes continue to evolve.
The evolution of IIoT technology building blocks started at the far edge, and continues today at the control layer. For instance, the emergence of multicore SoCs with integrated accelerators, and the adoption of networking standards like Ethernet TSN, have paved the way for improved device management and the implementation of containerised enterprise applications.
These innovations will undoubtedly continue to play a key role in defining the future of IIoT.
| Email: | [email protected] |
| www: | www.trxe.com |
| Articles: | More information and articles about TRX Electronics |
© Technews Publishing (Pty) Ltd | All Rights Reserved
printer friendly version