The Internet of Things (IoT) pushes measurement analytics to the edge of the network, redefining the sensor’s place in the electromechanical ecosystem. No longer a discrete component working in isolation, the sensor interacts with computing and communications components to provide intelligence via two-way communications. As a result, deploying sensors in the IoT presents different challenges than those encountered when integrating sensors into traditional network-edge devices, placing greater demands on traditional practices and requiring new design and test approaches.
To perform their functions, most IoT sensor nodes combine a sensor front-end, signal conditioning, power supply, microcontroller and back-end communications. For the engineer, the challenge is to ensure that not only do these components work well together, but they also work with other connected “things,” creating harmony from diversity.
To meet this challenge, engineers must adopt a systems engineering perspective, looking beyond individual components and viewing the sensor as part of a larger whole. Using this perspective, engineers must determine how the sensor fits and interacts with the other components in the node. This demands that mechanical, electronic and software engineers work together, viewing the system holistically to ensure the node meets performance specifications.
Fitting More Into Less
In addition to making all of the components work well together, design engineers must also fit more and more functionality and components into smaller spaces. These design demands place a premium on miniaturization and packaging technologies.
For example, engineers have traditionally considered microcontroller and wireless transceiver requirements separately, but for an IoT system, the selection of these two components must be considered together. Instead of using discrete components, designers now have the option of using a system on chip (SoC) components that combine transceivers and microcontrollers. These devices often include extra memory and processing power, saving space and reducing cost and power consumption.
Another option is micro-electro-mechanical systems (MEMS) technology, which has evolved from miniaturized, single-function systems into complex integrated systems. The latest generation of these systems brings together sensing, data processing and actuation, enabling them to handle a number of IoT functions.
The adoption of MEMS and SoC integration approaches can simplify the development process. They do, however, make new demands on design, prototyping and testing tools.
The Cost of Openness
Engineers must also ensure reliable communications. IoT edge nodes communicate over physical layer protocols, including a wide variety of wireless standards, ranging from Bluetooth to cellular, constantly pushing the limits of range and throughput. The challenge here is that no one standard meets the requirements of all applications. As a result, engineers have to select the technology that delivers the required levels of performance, coming to terms with interoperability hurdles and ambient interference. To meet these needs, they can turn to tools and platforms that analyze RF (radio frequency) signals and communication protocol exchanges.
The downside of wireless communications is that they open the nodes to security threats. One of the biggest challenges of providing security for the small endpoints of the IoT lies in the fact that traditional techniques of isolating the nodes often do not work, and hardware-based security cannot always be applied. Engineers do, however, have the option of turning to software-based security solutions.
Less is Better
The power consumption design criterion for IoT sensor nodes has one rule: Do more with less. This means that the node must be able to sense a physical property, perform analytics and transmit data to the Internet on a significantly reduced power budget, regardless of the power source technology.
This calls for hardware- and software-based power management. To a large extent, the solution lies in a combination of ultra-low-power optimized software that supports a variety of power-saving and sleep modes and energy-efficient hardware (such as point-of-load and fanless, conduction-cooled power supplies).
Many engineers designing IoT devices may not have a depth of knowledge in all the fields touched by the design process. As a result, designers will require intuitive, easy-to-use test and prototyping systems that provide a holistic development perspective and enable shorter design cycles.