Good to be back! I decided to become a terse writer: mostly because I am typing on a Surface Pro right now. Here are the updates:
– Thanks to Nick, upod-dev website is up!
– Thanks to Melwyn, HomeOS can log both with and without the initial handshake with Helios!
– Ordered sensors, devices, tablets, and laptops
– Multisensor is currently attached to one side of the lab room and collecting data at an angle like Sauron’s Eye.
– Multisensor views is up on upod-dev website!
The reason why it takes two resets to pair a zwave device to HomeOS seems to be because it takes two manual unpairing for zwave and multisensor to unfriend each other.
I have been manually unpairing zwave dongle and multisensor every time starting up HomeOS so that it can pair these two on its own at a fresh clean state.
Initially it threw this error:
But when I unpaired manually again, it didn’t have any port issues.
The multisensor named “double_negative_pair” is thus collecting data currently.
When I do not manually unpair, the devices are already connected and as soon as I start the script to get HomeOS running, the sensor that was previously connected will start to log.
This means that even if you add a new sensor with a different name, the logs will contain data from the previous multisensor.
Figured this out by giving multisensors some funky names.
Multisensor’s button is a bit off-center, so you need to press at a 45 degree angle to the right to get the button click registered.
Purple Surface Pro’s Visual Studio license has expired.
– Check Helios database to see if multisensor is currently only logging locally or sending to Helios
– Check correct parsing of multisensor data when sending to Helios
– Get Helios GUI on website set up so that we can easily check whether data is being logged on Helios without the need to log into Helios VM
-Develop auto-generation process of adding devices to Helios to make the process quicker
-Contact Anavi with specific to-do list
-Add door sensor
-Add noise-level detection in Kinect
-Add more devices
I personally am not too sure how useful door sensor data will be if they are installed on room doors, although implementing it will probably be (hopefully) easier then multisensors. Multisensors will detect motion and lights can tell whether it is turned on or not, so we can easily detect presence in room, and if we install several multisensors, we can detect movement flow as well. It is hard to tell which direction a person is headed — out of the room or inside — just from door opening and closing.
It could be useful if we install them on drawers, windows, cookie jar, trash can with lids, fridges, or toilet seats.
Actually now that I think of it, installing them on windows could be pretty useful, correlating them to temperature and humidity data.
Unless we can get the above IRB approved, I propose that we look into adding other sensors or use noise-level detection functionality of Kinect!
Tomorrow, Tushar will leave a mark on the blog and we will take turns on keeping this blog alive.
And please comment with compliment, criticism, random ideas, encouragement, discouragement, anything!