Sensirion's Wearable Development Kit, based on Pebble Time
While reference designs like the Atmel Xplained Pro, the Raspberry Pi, or Arduino offer a great way to get started developing products with our sensors, the form factor makes it difficult to build truely portable devices. At the same time, there has been a lot of interest in sensors for mobile products, from simple accessories over wearables to smart phones. In order to build prototypes for use cases in this space, Sensirion was looking into a number of options to provide a solid platform to build prototypes, algorithms and proof of concepts with little effort, removing challenging parts like battery power, charging, programming etc..
To make it easy for developers to get started, we eventually chose a solution that's based on the Pebble Time. The Pebble Time not only allows users to write their own apps, it also has a serial interface that allows third parties to build straps with additional functionality - for example to add sensors. In our case, we took it a step further: we build an additional housing (called "Backpack") where the Pebble clips in, and added some extra features: storage for logging, a microcontroller to run algorithms, and two of our temperature and humidity sensors. Our backpack is also Arduino compatible, so both the Pebble apps as well as algorithms for the Backpack can be written using existing toolchains that are widely available.
Finally, our engineers have ported some of our existing demos to the Wearable Development Kit, both for us to use at conferences and events as well as for you to have some inspiration to get started!
A Note on the Future of the Pebble Platform
You may have read that part of Pebble's software assets were acquired by Fitbit in December 2016. At this point, Fitbit is planning to keep the Smartwatch services running at least anohter year. More information on this can be found here.
At the time of writing this, there's still Pebble Time watches available for sale. We'll keep an eye out for a suitable replacement for the future. The demonstrator itself doesn't rely on Pebble's cloud services to work, so even if this was to go away in the future the demonstrator will continue to work fine.
AirTouch® is a touchless input method. Since the human breath has a very high humidity, we can detect when a user is blowing onto a device, and trigger an activity. The advantage of AirTouch® is that there's no movement of the device as it is the case when touching with a finger, which is great to take a photo without introducing blur. For wearables, it's a great way to scroll through content without obscuring the screen with the finger, or using gestures that move the device. It's also silent, so unlike voice commands it can be done in a quiet environment without causing attention or disturbing others.
Similar to the perspiration use case, we can use the two sensors to detect whether a device is worn or not. Typically, this is interesting for battery powered devices like headphones/earables or wristworn devices to enable functionality. This can be used to save power, but also to trigger actions: imagine headphones connected to a phone, with the music automatically stoping - or switching to the speaker - when the headphones are taken off. Learn more about on/off body detection by clicking here.
Since we have two sensors in the Wearable Development Kit, we can look at humidity differences between those two. This allows us to estimate the amount of water that we lose through perspiration. Applications in that space could warn about dehydration, or give us better information about workouts. Learn more about Perspiration by clicking here.
Thermal Context Awareness
A lot of research these days is covering context awareness, which allows devices to understand more about both their own context (device context) as well as the context of their users. This can be used either directly on the device to adjust certain functionality, or in services later to analyze user behaviour. Since our focus is environmental sensing, we started looking at situations where the use of temperature and humidity could help the user, especially when it is tied to the users perceived temperature. We refer to this as "feels like temperature". This is not necessarily the same as the (average) temperature in a room; for example two persons can be in the same room, but one just came back from a run. That person will most likely feel a lot warmer. Other factors that influence perceived temperature are things like direct sunlight, clothing etc..
Our demo app for the Wearable Development Kit classifies the perceived temperature into five classes, from cold to warm. This can be used as a basis for new apps. Some of the ideas we had are:
- Smart AC/heating control in apartments based on actual "feels like" temperature
- Remote care and assisted living for babies or elderly, where the device can warn early in case of discomfort
- Fitness tracking based on environmental conditions: lap time might be lower on a very hot day, but that does not mean it was a worse performance
- Sleep logging, along with automatic control of the AC or a heating mattress pad
- Include in social media posts ("Just came back from a run - feeling hot")
Build Your Own
We're have prepared all source code, schematics, gerber files and 3D printing information to allow interested parties to build their own Wearable Devkit. This information can be found in our "Building your own device" tutorial.