The smart home concept has been around for a while, but adoption has lagged behind the hype. Now, however, new enabling technologies are fostering a surge of IoT smart home innovation and use cases. Matter, AI at the sensor Edge, Wi-Fi audio, ultra-low power (ULP), direct-to-Cloud connectivity, and enhanced security are all available and are providing the impetus for both adoption and design innovation.
Technology advances are yielding untethered IoT smart home device solutions with unprecedented connectivity options and interoperability. Simultaneously, standards are now sufficiently advanced to promote rapid innovation in energy optimization and enhanced security. Artificial intelligence and machine learning are enabling new applications and enhanced features for existing ones. Now, homeowners have easy set-up of smart devices, and real-time monitoring and detection. Here are the top enabling technologies, and what they bring to smart homes:
Matter and Thread
Two of the most exciting advances in smart home technology are provided by the Matter Interoperability Standard and the Thread protocol. They enable IoT device manufacturers to deliver easy connectivity and interoperability with devices from multiple providers—long a goal of smart homeowners and IoT device manufacturers.
Matter 1.2 is the next round of updates to the Matter protocol specifications, introducing new device types as well as interoperability and user experience improvements. With Matter 1.2, the list of solutions includes white goods, fans, robot vacuums, air quality sensors, smoke alarms, room air conditioners, and more. Matter 1.2 also adds features that improve commissioning that better describes the product or its elements. Thread allows devices to communicate with each other directly, but unlike Matter over Wi-Fi, it needs a router to connect Thread devices to Wi-Fi or Ethernet for Cloud connectivity.
The combination of Matter and Wi-Fi ensures users will have a more seamless experience across various product types and applications. Matter enables devices to communicate over Wi-Fi to an Amazon Echo, Google Home Pod, Apple Home Pod, Samsung, etc., allowing the consumers to interoperate their Wi-Fi devices in the home with devices and controllers from these and other ecosystems.
The rapid advancement of artificial intelligence in recent years has also opened opportunities in Matter-enabled smart homes.
AI at the sensor Edge
When AI computing occurs at the Edge, data does not need to be sent over the network to be processed on the Cloud, nor is Internet connection reliance necessary. Instead, algorithms use data and computing power generated by the device. Computations take place close to where data is collected rather than a centralized Cloud computing facility or remote data center. By bringing AI computing to Edge devices, the processing of deep learning algorithms takes place locally right at the sensor Edge.
These algorithms handle computations, run inference, and make independent decisions in near real-time. Inference, for example, occurs where sensor data id is generated, enabling low power and low latency. Increasingly Edge AI use cases require high-speed processing with minimal latency yet high accuracy. Latency reduction means users can operate smart speakers with voice commands in real-time.
Akin to Edge AI is Tiny Machine Learning, also known as TinyML, which performs machine learning inference on smaller, performance and power-constrained devices and embedded systems. TinyML models, including machine monitoring, eliminate the need for large language models currently running only in the Cloud.
Sensor Edge AI computing is enabling new applications and use cases requiring continuous operation, fast decision-making, quick responses, and low energy. Sensor Edge AI saves money previously spent on bandwidth, Cloud processing, and transferring video, voice, and sensor data over the network. It also allows the use of AI models on always-on use cases such as motion/gesture detection, audio/person detection, and others that must respond immediately.
Wi-Fi audio
Many smart device functions rely on Wi-Fi, including battery-powered video doorbells that upload video recordings to the Cloud. Streaming lossless audio requires throughput higher than Bluetooth Classic or Bluetooth Low Energy (BLE). Streaming lossless audio over a cellular or Wi-Fi network consumes significantly more data, and downloading lossless audio requires more space on your device. Higher resolutions use more data than lower ones.
Lossless audio over wireless headphones is somewhat complex. For example, given Bluetooth’s 2,000kbps bandwidth limitation, most wireless headphones don’t support complete lossless formats. The devices that include ear pods and wireless speakers must be battery-operated, and Wi-Fi audio streaming must be performed at very low power. Streaming from a Cloud source to a headset allows consumers to roam anywhere within the smart home when connected to the access point. An MCU is integrated into the Wi-Fi SoC to support Android or Apple ecosystems.
The power behind new IoT devices
Ultra-low power (ULP) is achieved using low throughput wireless technologies such as BLE, Zigbee, sub gigahertz, none of which offer direct Cloud connectivity. ULP is necessary at a chip/module level, and at a complete device level, including power efficient end-to-end (E2E) connectivity to the access point/router and the Cloud.
Ultra-low-power leads to longer battery life, smaller battery size, and miniaturized IoT device form-factors. IoT devices featuring ULP Wi-Fi capabilities such as BLE power levels are critical for battery-powered devices to achieve a 2-4X enhancement in battery lifetime.
For several smart home applications, especially security, most of us have perched precariously on a step ladder every 3-6 months to change out batteries in our outdoor cameras. By achieving a long battery life (greater than one year), we not only reduce the opportunity for mishap, but also use more environmentally and sustainable solutions.
Today’s designers make every attempt to design long battery life devices for easy Cloud connectivity that interoperates with multiple platform ecosystems. More recent smart home use cases, such as appliances, require a battery life of between 5-10 years.
Direct-to-Cloud connectivity
BLE and Zigbee devices today must connect to a gateway device that provides Wi-Fi/Ethernet connectivity to the Cloud. A bridge/router use can be avoided if devices have Wi-Fi integrated to directly connect using an existing access point in the home. This direct-to-Cloud connectivity delivers ease of use and installation. Standards-compliant Wi-Fi is interoperable with hundreds of access points deployed within the smart home, enabling easy direct-to-Cloud connectivity to IoT devices.
Enhanced security
In the early days of IoT, it was not uncommon to design and manufacture a device and then fix the security aspects of it later. In today’s world, however, unsecured, or poorly secured IoT devices risk being compromised and provide an entry point to hackers, making the complete smart home network vulnerable.
Security technologies include secure boot to alert users should the BIOS be compromised, an on-chip encryption accelerator that enables data transfer to the Cloud, and support secure protocols such as SSL/HTTPS, etc. The use of e-fuse PUF technologies also provides unique device keys that help secure IoT devices.
Coming together…
In addition to the sum of the parts described above, there’s another important element aiding OEMs and designers, especially those that are not necessarily familiar with each of the new technologies.
Although Systems on Chips (SoCs) make sense for designers who are extremely familiar with RF design and designs that consume a very high volume of SoCs, for most IoT designers, access to a fully certified and compliance-tested wireless module solution makes their job easier and faster. When modules are tested and certified by the manufacturer, there’s no need for re-certification of the complete device.
Individually, the above-mentioned technologies and standards advancements are a boon to IoT development. Their simultaneous availability today, however, goes well beyond that boon—enabling a myriad of use cases, while consistently breaking old barriers to development. What’s critical now, however, is having the right supplier/partner to avoid mistakes and delays in such a fast-moving segment.
Deepal Mehta is Senior Director of Global Marketing and Sales at InnoPhase IoT.