Genie Community Forum

ThingPedia devices that monitor physical parameters

I am designing a demo for a system that uses Web Almond (amongst other things!) as a source of information. To even get started I need a ThingPedia device that monitors something. It can be anything that is a physically measurable quantity (temperature, humidity, a smoke alarm, etc), at least to get me going. Every such device I have tried from the MyAlmond $\to$ Configure New Device page has failed when I tried to add it. In fact, most such devices have been removed from the page.

Have I done something wrong? Is there some constraint that I am not aware of? Please advise…

Hi!

The best way to enable a device that has some measurement capability is through Home Assistant.
First of all, are you trying to use Web Almond (aka almond-cloud, aka almond.stanford.edu), or almond-server (the addon embedded in Home Assistant)?
You should follow the instructions at https://www.home-assistant.io/integrations/almond/ to configure Home Assistant with Almond. Or if you use the addon, everything should come preconfigured and you should have your devices available.
Once Home Assistant is configured, all devices known to Home Assistant will be visible to Almond. If some device is not shown, it might lack the appropriate “device class” information in Home Assistant, or it might be of an unsupported type.

Other than Home Assistant, you can also monitor pretty much anything in Thingpedia. For example you can monitor the weather (“notify me when the temperature in san francisco goes below 50 degrees”).

Hi Giovanni,
Thank you very much for your prompt response and advice. I am not an AI researcher. I am just a simple physicist and sometime software developer. I am designing a prototype for a start up, and I am thinking of having my prototype use Almond as an information resource.
I don’t think Home Assistant is the approach we want to use, since we don’t need to monitor static local sensors. Rather, we need to collect data from anywhere on an as-needed basis. One scenario might be something like this:

  1. We, for various reasons and in various ways, receive a set geolocation coordinates.
  2. We go to an IoT search engine like Thingful.net, and search for sensors that collect the sort of data we need.
  3. We identify the particular sensor that we will use.
  4. We go to Almond and request the data from that sensor.

Of course, we would use Almond for many other things, but to get started I would like to choose a simple scenario with simple data, and implement it so that I can learn the nuts and bolts of how to connect and request and receive data. Does that make sense?

I have set up a temperature monitor on Almond using a local weather site, but now I would like to try something closer to our real application. So I guess my questions are 1) Are there simple IoT sensors that I can read out through Almond from my own little application running from work, and if so, 2) what sensor should I use? I don’t really care what the sensor is measuring or where it is, to get started.
.

We don’t really have IoT sensors that you can connect directly to Almond right now, our IoT support comes through Home Assistant.
At the same time, you can try adding a new sensor yourself. Thingpedia is an open repository, and we have documentation and tutorials at https://almond.stanford.edu/doc/thingpedia-intro.md
You can add a polling sensor, or use some other protocol, such as WebSocket.

Also, checkout our Thingpedia skill implementations at https://github.com/stanford-oval/thingpedia-common-devices for inspiration and examples.

Excellent. Thank you very much, this is exactly what I needed. I will try to make some individual device work and then report what happens. Thanks again for your help.

Rowan.

The availability of wearable devices (WDs) to collect biometric information and their use during activities of daily living is significantly increasing in the general population. These small electronic devices, which record fitness and health-related outcomes, have been broadly utilized in industries such as medicine, healthcare,and fitness.
Since they are simple to use and progressively cheaper, they have also been used for numerous research purposes. However, despite their increasing popularity, most of these WDs do not accurately measure the proclaimed outcomes.
In fact, research is equivocal about whether they are valid and reliable methods to specifically evaluate physical activity and health-related outcomes in older adults, since they are mostly designed and produced considering younger subjects’ physical and mental characteristics. Additionally, their constant evolution through continuous upgrades and redesigned versions, suggests the need for constant up-to-date reviews and research. but now I would like to try something closer to other platform to remain busy.
Accordingly, this article aims to scrutinize the state-of-the-art scientific evidence about the usefulness of WDs, specifically on older adults, to monitor physical activity and health-related outcomes.
This critical review not only aims to inform older consumers but also aid researchers in study design when selecting physical activity and healthcare monitoring devices for elderly people.