Hi all,
It’s been a while since I’ve been active on the chatbot forums lately. There’s an excellent reason for that: I’ve recently joined a new internet startup: all things talk. We are building the next generation of intelligent, cloud based home & business automation systems.  I’m currently working on the AI & UI parts: All very exciting stuff.
The first version probably wont have any chatbot features, but that’s just a matter of prioritizing. As my skills are more centered around software development compared to true chatbot-masters, there might be an opportunity there in the future, when those parts will be added.
Anyways, it’s full throttle development time for me at the moment, so my online alter ago will probably remain fairly dormant for the next few months. When the first (beta) version is live, I’ll probably start to blog some more about tips and tricks on how to automate your home…

Nice, I wondered where you had gone Jan   Sounds interesting, good luck with that

Thanks. yep, very interesting indeed. Also challenging…

Hi Jan. I, too, wondered where you were.

Can you give us an idea of the sort of applications that are being planned.  I keep hearing (or seeing) stories about how our refrigerators, microwave ovens, and dishwashers will converse with us one day.  To be honest, I really don’t want a car that chats or a cell phone that tells jokes, and I’m already at my limit as far as that goes… I really don’t want to have a conversation with appliances as I try crossing through the kitchen, or to be slapped with advertisements tailored to me personally as I pass each store at the mall.

The initial idea is not so much to have your machines talk to you, but each other. There are currently already a few systems out there that allow you to automate things : zigbee,z-wave, x-10,...  but these protocols don’t talk to each other and they can’t use data from the internet (weather api,...). Often, these devices are also low powered and can’t directly connect to the internet (for which we are developing low-cost gateways, probably based on raspberry-pi). What we are currently building is a layer on top of these already existing protocols so that they can all work together: use an RFID reader to start/stop a zigbee device for instance. Another scenario could be: You are on holidays when the cleaning lady (or a famiily member) calls to tell you they forgot their cellphone (or laptop or whatever) at your house last weekend, which they urgently need and if they could go and pick it up. You could then use the cloud-app to let the door give temporary access to that person. It’s also a great way for small businesses to do low cost automations.
My job is to add those virtual devices (web-apis) and build a nice looking UI that allows you to draw diagrams for building all these different scenarios. Some sort of ‘recording’ (or learning) abilities will also be possible. And for the die-hard coders, you’ll probably be able to write those diagrams as scripts.
Verbal commands might be for a next version (not part of the initial release) although it will probably be possible from the start to have a fridge (or better, a sensor within the fridge) that is able to email you to inform that something is starting to stink.

Fascinating project. Sounds a bit like adding Bluetooth to all devices. I imagine it would be of use to businesses or people who live by routines, but it raises the question of how much control you want over your appliances, and how much you want them to do on their own.
For instance, my morning routine may include having a coffee 5 minutes after I hit snooze on my alarm clock, but I may one morning not feel like having coffee. If my coffee machine activates as per the usual morning routine, that would be a waste of coffee. So instead of having to activate devices myself, I may find myself trying to keep in mind which of all devices to deactivate, or be run by my devices. On the other hand I would find it most convenient if something would turn off my lights, electricity and heating when I close the door behind me. If only there was a button for that.

@Don, re: coffee, that’s where Speech Recognition and voice commands would be useful to have integrated into the HA system. Just hit that snooze bar and croak “COFFEE!” as you lie back for that oh, so pleasant extra 9 minutes of time in bed. To forego the morning java, simply omit the command.

“Tea. Earl Grey. Hot.”
What would be really cool is if I could just snap my fingers at every device to turn it off or on.

@Don: you are right about routines and people not always sticking to them. Which is why it’s probably best to allow the users define things for themselves: do you want the completely automated process or still have some control over it.
You could for instance define a button on your android’s virtual remote app to start the coffee bot when pressed. Or have the coffee pot started when a sensor in your bedroom gets triggered in the morning.

@Dave: I’m not certain if giving voice-commands will be in the first release or not, but it will definitely become a part of the system, either natively or through a plug-in system (feature will be present from the start) to allow 3th party developers to add custom functionality, like chatbot engines.

Don Patrick - Jul 12, 2013:

“Tea. Earl Grey. Hot.”

Is that the ghost of Cpt. Jean Luc Picard I hear?

For me it would be “Tea. Green. Tepid. With lemon” (can’t stand overly-cold beverages, first thing in the morning)

Sounds a very exciting project Jan. Good luck with it all.

Welcome back!!!!! Love to hear more about your project. A video would definitely be of help!
Cheers

Hi Erwin. It’s still a little early for videos. We are currently putting down the foundations of the platform, so there are only presentation slides with pictures. There should be a demo ready by September (probably with videos), with a first beta release scheduled round April next year (very optimistic timing).