Welcome to the IoT weekly round-up. This week, we’re patting ourselves on the back as the Watson IoT blog pockets Platinum and Gold dotCOMM awards. In other news, there are new connected devices for plant and bee-lovers, and lost-item-tracker company Tile unveils a new line of products with better range.
Watson IoT blog wins big in AMCP dotCOMM Awards
The Watson IoT blog is officially award-winning! We swiped Platinum and Gold dotCOMM awards last week from the Association of Marketing and Communication Professionals, and are rather proud of ourselves, thank you very much. We’d like to say a huge thank you to our stellar team of contributors, and to you, our readers, without whom the blog would be as useless as a chocolate teapot.
Internet of Bees hive health tracker
In an effort to keep a close eye on the declining bee population, Oldooz Pooyanfar, a Canadian researcher, is working on a connected hive monitoring system. It uses microphones, temperature and humidity sensors to offer a drones-eye view into a colony, and will hopefully be able to give some insight into colony collapse disorder, the causes of which remain somewhat mysterious.
HelloPlant keeps an eye on your garden
If your plants frequently fail to flourish, especially if that’s because you forget to water them, HelloPlant can help. It’s a tiny tool whose job is to monitor your plant from within its pot, and give you a heads-up when things are less than OK. HelloPlant can keep you abreast of soil moisture and light levels, and modifies its notifications according to plant type, so you always get the right info.
IBM announces faster distributed training time for visual recognition models
Without wanting to toot our own horn too much, we can tell you that IBM has some pretty impressive numbers to offer this week where AI training methods are concerned. IBM’s research group was able to train the ResNet-50 image recognition model for 1k classes in 50 minutes across 256 GPUs. If that means nothing to you, you just need to know that it’s faster than the training times recently achieved by Facebook’s AI Research Lab, which were none too shabby either. What this means in practical terms is that IBM is working towards breaking down deep learning problems into bitesize chunks – in order to achieve faster model training for customers.
Welltory app measures your heartbeat to determine stress levels
You know what would help when you’re feeling stressed? Knowing exactly how stressed you are. Usually you’d need some serious hardware to figure that out, but that might be about to change. A New York startup has developed a method that uses heartbeat readings taken with a smartphone app, which algorithms and machine learning capabilities then analyse to determine stress levels. The idea behind the app, which goes by the name Welltory, is to measure the effect of various lifestyle factors such as diet, exercise and meditation on stress levels, so that users can identify helpful habits and discard the rest.
Tile launches lost item tracker Pro series
We’ve written about Tile lost item trackers before, and it seems that they’re building on previous success with a new line of premium devices. The Tile Pro series trackers sport an upgrade that means they can be located from much further away. Externally, there have been improvements too – the new devices are better at repelling water, have a louder ring, and are made from tougher and more flexible materials. The new series comes on the back of a $ 25 million funding round.
Keep up-to-date with the connected world
Bookmark the IoT weekly round-up series page to keep up with what’s going on in the wider world of IoT.