Designing a Skill
While games and trivia are fun, I think the real-life use model for ambient computing devices will be retrieving information whose usefulness has a short shelf-life.
My Skill
That said, I always find it to be a pain to track down security line wait times at the airport–that’s going to be my skill. Normally I could care less about TSA wait times, but when I need the information it’s always a gigantic PITA to track down the website and jump through the user inteface hoops to get the information. I was surprised this infomation is published in XML format from the TSA, it’s published consistenly, and without a ton of credentialling to access it.
Voice Interaction Model
Intent Schema
I think I’ll have three custom intents:
getWaitTimeByAirportCode
getWaitTimeByAirportName
getWaitTimeByCityAndState
and three vanilla intents like:
AMAZON.HelpIntent
AMAZON.StopIntent
AMAZON.CancelIntent
Sample Utterances
My sample utterances will preliminarily consist of the following:
- “What is the wait time for [airportCode]?”
- “What is the wait time for [airportName]?”
- “What is the wait time for [airportCity], [airportState]?”
Visual Representation of the Menu/Model
Below is a “physics for poets” representation of what the the model will look like for retrieving information. Based on the XML files I’ve looked at, there will be a two special cases to address, such as:
- cities with multiple airports and
- airports with multiple security checkpoints.
Companion App Cards
I’m not going to include companion cards with my app, as my primary objective is to not to manually interact with a physical device.