Spark Core Bathroom Monitor | Hacker News Reader

Spark Core Bathroom Monitor

americastestkitchen.github.io

42 pointsnixpulvis15 comments

Wow, this just makes me love America's Test Kitchen even harder.

Edit: and it's served from a statically-generated jekyll blog that is on their github page. Did I mention they have a github account. Very random and awesome.

In case you haven't seen it, here's the long-running Random Hall bathroom server: http://bathroom.mit.edu/

Hadn't seen that before, very cool. What do the "*"s mean next to the bathrooms.

Oh I'm an idiot, they are just a easier way to see occupied.

Does this method fail if someone closes the door behind them when they leave the bathroom? I'm not sure of the exact setup of the bathroom, but I can imagine that many people would instinctively close the door behind them rather than leave it open. Maybe at ATK they have a strong culture of open = unoccupied.

What would be ideal is if it could be wired to the locking mechanism of the door. That's probably decently simple to do if the door is locked using a deadbolt because something physical is actually entering the doorframe. If it's just a lock that prevents the doorknob from turning, then you'd likely need to disassemble the locking mechanism and see where you can attach a switch.

There is a strong culture for leaving the door open, to avoid the awkward door knock. This defiantly does mean we can have false "occupied" states, but in practice at our office this doesn't happen.

Actually, if you were to implement something that monitors the state of the door lock, the same hardware could run an OCCUPIED/VACANT LED on the door itself to avoid the awkward door knock in the event that the door is closed while the bathroom is unoccupied.

Of course, if someone forgets to lock the door, that's basically an invitation for everyone else to walk in on them. On second thought, it seems to me that the best solution here is to mount webcams in the bathroom so that anyone can see if it's occupied. Seems like the only foolproof way to go here.

On a related note, we have a take home fridge at ATK, there was some discussion about how to tell when new food was added to the fridge. Cameras might have been an option there.

I would attach a motion switch to check whether any motion was detected after the last close. If so, someone is in.

My favorite part: "The next step involves complex analysis of historical bathroom visit data, to allow for statistical insight into bathroom habits. Oh, and a 💩 icon for when the bathroom has been occupied for more than 5 minutes."

Brings new meaning to the "The Quantified Self"

I wonder when the productivity benefit to ATK employees will outweigh the $$ they spent paying you to make this.

It actually would have been very interesting to poll employees and get data on 'failed bathroom trips' per day per employee before you finished the project.

From there, you could extrapolate the (supposed) daily productivity increase and do things like plotting the $$ or minutes saved per day on the screen.

Well aside from the fact that most of this development was done in my free time, I can say personally it's saved my a good amount of frustration and time. Hard to really quantify though.

Actually it's quite easy to quantify!

First you need to wire up switches in everyone's seats...

Here's another bathroom monitor from thoughtbot a while back: http://robots.thoughtbot.com/arduino-bathroom-occupancy-dete...

Reply on news.ycombinator.com