What do you see as the key challenges in sensors as they play a larger role in the Internet of Things (IoT)?
For many businesses, the most important challenge facing IoT smart sensors and the Industry 4.0 revolution will be the data generated by billions of new devices. Cisco estimates that sensors will generate 507 zettabytes of data in 2019. Managing and analyzing so much information is a massive challenge, and instead of sending all of it back to the cloud or a remote center, many companies are reverting to an edge-computing model--storing the data locally on industrial-grade solid-state drives and applying machine-learning algorithms to extract the actionable information.
The consequences of a lax data-storage approach to IIoT systems can be grave; a factory's entire operation or a smart city's grid, for example, could be compromised. To combat such outcomes, system designers must observe these imperatives early in the design process:
1) Critical data should be stored locally in case of network interruptions, ensuring data can still be acted upon until the network is back up.
2) Data security involves securing data at rest and authenticating systems capturing sensor data, such as machine-to-machine authentication, so that a threat acting as a trusted sensor can't invade the system.
3) Sensor warnings in critical applications such as robotic surgery, collision avoidance, and cockpit malfunctions require real-time action, so low-latency in storage is essential.
4) Not all sensor data is useful; statistics show that less than 20 percent is. So, incorporating data analytics at the edge to filter meaningful data before being stored can save time and money.
By Jennifer Barry, product line manager, Sensing, Texas Instruments
As IoT continues to broaden its reach across applications, so does the need for multiple sensors. Sensors are used for various reasons, ranging from position to environmental to protection of the system. One of the biggest challenges facing designers today where multiple sensors are integrated is fast response time.
Every consumer has his or her own expectations of how an end product should work. Typically, this consists of convenient, quick, and reliable. Convenient translates to data anytime you want it. Quick implies immediate. Reliable states that it works every time with no disruption. These are high expectations, and not always easy to meet when multiple sensors are required in IoT systems. Response time plays a key role in meeting these expectations.
Response time is important to the sensor, data storage, and controller in the system. When looking at the complete system, the control loop of communication between these is crucial. Sensors collect data in a customized manner, either in continuous mode or triggered by an action that occurs. Once the data is sampled, it must be stored in memory. Slow response time could result in missed data from the sensor. In some applications, missed data can be minor and seen as a glitch to the consumer. In other applications, missed data could have a larger impact.
Engineers working on IoT projects have challenges with an increased number of components on the board--many of which are becoming sensors. As sensors transition to a key component of the IoT system, response time will continue to be a key element requiring innovation.
What do you consider to be the major trends in RF and microwave components?
By Harvey Espinoza, director of product marketing.
Communication Devices Business Group, TDK Corporation of America
Filtering requirements have not only multiplied, but have also become more complex due to the ever increasing RF bands for 5G, new services, and WiFi expansion. For example, adding the n77 (3.3-4.2 GHz) and n79 (4.4-5 GHz) 5G bands to a device requires new filters for these bands as well as updated filtering/rejection requirements for existing filters that cover other cellular bands and connectivity (WiFi, BT, GPS, etc.). Many of these new bands are location specific, such as the n79 band right now being planned for China and Japan. Perhaps some devices do not have to support these regional bands, which will make their filtering requirements easier. But if you want to develop a product with worldwide support, this becomes a challenge.
The close proximity of bands (sometimes 100 MHz apart) also requires new filtering technologies. To meet all these challenges, there is not a single solution. F-BAR and BAW/SAW filters can do some of the best filtering, but with a price premium. Ceramic and thin-film filters can do some filtering, but the low cost does come with a performance limitation. Hybrid designs that combine two different technologies are under development to meet some of these challenges, with the vast majority of hybrid filtering solutions still to come. The hybrid filter can be in one package or part of a solution. All of this filtering will need to be developed while keeping the same size or downsizing, and with high rejection and low IL. This filtering challenge will keep RF engineers busy for years.
|Printer friendly Cite/link Email Feedback|
|Publication:||ECN-Electronic Component News|
|Date:||Jun 1, 2019|
|Previous Article:||OKW Enclosures, Inc.: With a mission of "To each his own housing," this enclosure manufacturer seeks to offer individualized solutions with a simple...|
|Next Article:||New & Noteworthy: Compiled by the ECN Editorial Staff.|