When you live in a sprawled suburb like the Silicon Valley, it’s sometimes hard to predict how long it will take you to drive from point A to point B. For example, my one-way commute time from Mountain View to San Francisco has ranged between 45 and 120 minutes in the past few months.
Of course the travel times are somewhat predictable. I know that if it’s a nice day and I leave SF at 10 am, the odds of time(travel) < 50 are high. Maybe once a week or so there is an accident that sets me back 10-15 minutes. Friday afternoons are different, coming back from Mountain View usually takes patience; I know that if I leave around 6 pm I probably be in the city by 7:30.
The obvious question is, what if I need to be in Oakland for a 5:30 pm meeting for which I just cannot be late? Google says it would take me 50 minutes if I left now. It’s 3:55 as I write this, if I were driving away this second I would arrive at my destination by 4:45. However, it’s raining outside. What are the odds that there will be an accident somewhere along the way? How much does traffic fluctuate during this time of the day, on a Tuesday? The point is that when I have an important appointment tens of miles away I have to make sure to pay attention to the clock since much earlier, which is distracting. What I would like is for Google (or whoever) to tell me when I need to leave A order to be at B before time T. Also, I’d like two modes:
- 80% confidence: I probably will be on time, but if I’m late it’s no biggie.
- 99% plus: most likely I will be early, but a high chance of being on time is worth the extra 20 minutes or whatever.
This could be done analyzing historical data, real-time traffic, and weather conditions. I would want to run the query in the morning, and have the service send me an update (SMS, email) any time my suggested departure time changes. Would I pay for this? Possibly, but I’d have to try it first.