February 2013 // 2013-55

Thinking About the Near Future: Scheduling

In the near future, my calendar will be intelligent.

Scheduling and tasks will become more predictive and responsive, and the program(s) will adapt and learn based on my preferences and feedback. I will use an app similar to Cue, but it will be more personalized and sophisticated.

As Cue already does, this app will take note of my daily weather/forecast and sunrise/sunset. Unlike current digital scheduling assistants, it will provide recommendations about what to wear when I step outside. Apps like Dark Sky already gather micro weather data and provide alerts just before it rains. Apps like Swackett also offer (somewhat cheeky) weather-related fashion advice.

My future scheduling app will know that I prefer to run most errands during daylight hours, but that I try to avoid the noon lunch rush. If I input my to-do list, it will recommend the best times during the day when I should go out to the grocery store, the post office, and the bank.

With micro weather data, it will also be able to shuffle these errands around based on the forecast. Icy conditions in the morning? Push the driving until the afternoon. Big storm coming in 3 hours? Tell me to go out now.

Speaking of errands, this future platform will learn my local behavior and patterns based on all of the location data that I have built up. It will know which grocery stores I frequent, where my mailbox and bank are, which library I like to swing by, and the routes I tend to take. It will keep track of the daily hours for these primary establishments, and it will be able to pull in current traffic data and make predictions, just like Google Maps.

Currently, Cue (which I’ve been turning to more and more as a calendar reference) pulls in the info from order tracking emails I get, and it shows on my daily calendar when a UPS or FedEx package is going to arrive. My future scheduling app will consider this info and reserve time for me to visit the mailbox. It will also know when I typically receive paychecks from clients (if I’m still receiving old fashioned paper checks in this near-future world), and it will consider the probability of picking up a check or other important mail item when planning a trip to the mailbox. It will know that I tend to withdraw about $60 in cash every two weeks from the bank and (if I’m also still using paper money in the near future) it will schedule time for me to swing by the bank after the mail.

My future scheduling app will know that when I’m doing a long errand run, I typically hit a grocery store, a gas station, the mailbox, the library, and the bank - in that order. If I’m doing short errands, it’s just the mail and the bank.

My library book pickup notices get sent to my email, so my scheduling platform should be able to pull in those details and add the library to the errand list when appropriate. In this near future world, it might even be able to communicate with the computer in my car, noting when I’ve hit 1/4 tank of gas and adding in time to visit the gas station.

I try not to schedule too many meetings, but this platform will list my commitments for me each day. I’ll learn to set time blocks for specific tasks, and the app will learn my work behaviors. If someone emails to schedule or reschedule a meeting, the app can respond and accept it or propose a different time. It will learn that general work tasks can be scheduled over and rearranged to accommodate for meetings and appointments, but I’ll need to start setting priority time on my calendar that can’t be double booked or bumped (for lunch, important uninterrupted work, etc).

If I have meetings or outside appointments, the platform will be aware of the location, and it will have directions ready to go if needed. It will know my preferred public transportation options, and will query the time schedules to tell me which light rail to jump on and when. If I have to drive, it might be able to track down a parking spot for me in almost real time.

With home sensors abounding in my hypothetical near future living space, the platform could log when the cat sitter (or dog walker?) visits during the day, when cleaning people show up, or when kids get home. If I’m traveling, it will be able to query my email and present me with my travel times, reservation numbers, and flight delay alerts. A variety of apps already do this.

The hardest change for me initially will be the act of recording more of the structure for my day so that the platform has something to learn from. I already use time tracking for work; maybe this will also help my focus and productivity. I currently log hours for work tasks after they’re done. I’d need to learn to set specific times for these work tasks each day. A lot of this becomes repetitive, so hopefully the platform will pick up on my patterns quickly.

This is a learning algorithm, so it will need feedback. Did you do this today? Yes. Did you go here? No. Was this a good time to run to the bank? Yes, but there was traffic.

With the data I already collect, a profile of my typical behaviors could be established fairly easily. Much of the information I’ve listed above is already available in a variety of current applications and platforms. This near future app could tie that information together and wrap it with a strong personalized learning algorithm. Right now I look forward to the near future and to the one AI scheduling assistant that will incorporate all of this information and begin making some of these plans and decisions on my behalf.

Present Calendar Apps

As mentioned, I turn to Cue a lot right now.
Other options include Sunrise, Tempo, and even things like Mailbox.



Author: Erin Jo Richey
URL: http://erinjorichey.com/thumos/articles/near-future-scheduling

« Older Post | Newer Post »

blog comments powered by Disqus