Supplement: Semiconductors
costly, exposed that data to unnecessary privacy risks, and compromised battery- powered operation. It soon became clear that the solution was to fi lter out the key, relevant data and make any data-driven decisions “at the edge”. Such edge computing enabled all kinds of advanced and exciting battery powered IoT applications.
A hot, cold and wet example One example is predictive maintenance and monitoring in fields such as the electricity power grid. Here utilities need to know as quickly as possible about critical events such as trees falling on transmission lines or fires, excessive winds, and ice. A TinyML solution could be used locally to detect faults on the power lines caused by such environmental hazards.
The challenge, even up a utility pole, is to get power to the sensors (high voltage overhead cables cannot easily be tapped for low voltage power). The solution is a battery- powered NB-IoT unit. One that if based on one of Nordic Semiconductor’s nRF91 Series low power SiPs, for example, could offer a decade or more of battery life.
The utility can then see the grid electric
signal in near real time and detect errors from things like lightning strikes, using a TinyML model, and then use the NB-IoT connectivity to send a signifi cant event notifi cation to the grid operator. Better yet, this data can also be collected to continuously train and optimise the underlying ML models.
Where’s the ‘killer AI app’ in IoT? AIoT is also about making familiar products better using historical data to make them more responsive and more effi cient. One example is a consumer refrigerator. The prosaic everyday white box in the kitchen, of which there are billions installed around the world, account for around 10-to-12 per cent of global electricity consumption. But what if you can put an ML cellular IoT device into the motor controller to monitor and adapt to usage patterns so that the refrigerator never has to work harder than it needs to? Then you would have a signifi cant impact on global energy usage. And, as another example, consider consumer washing machines. Again, there are billions of them installed around the world. And again, people rely on the machines so need them to operate effi ciently.
Nordic’s nRF9160 low power SiP supports Edge Impulse’s TinyML for local machine learning operation while offering long battery life
What if you could use AI and ML to monitor the washing machine for early signs of failure from simply monitoring the telltale vibrations a machine makes when operating abnormally? All washing machine manufacturers know what such vibrations occur as the machine wears. They just haven’t had the ability to monitor them in a technologically and commercially viable way before now. But for all the talk of AIoT and ML, the IoT revolution remains a somewhat ‘invisible’ one. It’s happening all around us in countless applications, but there haven’t
been one or more killer applications to really raise its profi le. Especially among the public. Is there such a killer app to make the market jump? That’s diffi cult to predict. But developers can use AIoT and ML to help solve some of the world’s big problems such as climate change, energy effi ciency, healthcare, social care, and economic parity. Then, although there might not be a killer app, the IoT will still be helping make the world a better, safer, more effi cient, place. And that is no bad thing.
https://www.nordicsemi.com/
EASY ASSEMBLY ‘ENCLOSURE PRODUCT OF THE YEAR’ WINNER DUAL-ACTION LID: Screwless push-button or one screw assembly
OPTIMISED SPACE: Large, unobstructed PCB footprint PCB FLEXIBILITY: Mounting bosses in base and lid
sales@camdenboss.com
www.cieonline.co.uk
+44 (0)1638 716101 Components in Electronics May 2024 45
Page 1 |
Page 2 |
Page 3 |
Page 4 |
Page 5 |
Page 6 |
Page 7 |
Page 8 |
Page 9 |
Page 10 |
Page 11 |
Page 12 |
Page 13 |
Page 14 |
Page 15 |
Page 16 |
Page 17 |
Page 18 |
Page 19 |
Page 20 |
Page 21 |
Page 22 |
Page 23 |
Page 24 |
Page 25 |
Page 26 |
Page 27 |
Page 28 |
Page 29 |
Page 30 |
Page 31 |
Page 32 |
Page 33 |
Page 34 |
Page 35 |
Page 36 |
Page 37 |
Page 38 |
Page 39 |
Page 40 |
Page 41 |
Page 42 |
Page 43 |
Page 44 |
Page 45 |
Page 46 |
Page 47 |
Page 48 |
Page 49 |
Page 50 |
Page 51 |
Page 52 |
Page 53 |
Page 54 |
Page 55 |
Page 56 |
Page 57 |
Page 58 |
Page 59 |
Page 60 |
Page 61 |
Page 62 |
Page 63 |
Page 64
