The idea for these graphs emerged from my universally shared experience of being stuck in traffic. Living in north Austin and commuting to downtown requires far too much of my life to be spent on I-35 and as a result I wanted to find a way to reduce my overall commute time. Having a reasonably flexible work schedule gives me the opportunity to adjust what time I leave and theoretically maintain the same overall working hours while reducing the amount of time that would otherwise be spent stuck in traffic.
In an ideal world, assuming flexible working hours, commutes would be done at minimal traffic times such that ~8 hours are spent at work with the minimal amount of time spent in traffic. This obviously would need to have reasonable constraints, driving into work at 11pm and leaving at 7am will most likely always have the least amount of traffic but is impractical for my job. With typical working hours being 9am-5pm, my expectations were for traffic to peak at ~8:30am on the way to work and at ~5:15pm on the drive home. Collected data assumes that the fastest route will always be taken and is adjusted relative to my fastest commute time. For example, if the fastest commute time with no traffic was 20 minutes, a commute of 40 minutes with traffic would show there being a 20 minute peak. If my standard route faced a closure and the new route given takes 35 minutes and has no traffic, the peak would be 15 minutes.
Data are collected using python and Google Maps API with the destination being between my house and work. Data are then stored on a MariaDB server, fetched with Node.js/Express.js, and graphed using D3 ridgeline graphs. Collecting the data gives me an overall solid idea for when to leave for work and provided me with real life and practical data to practice data analysis with. They also serve as a general insight into what Austin traffic is like as the majority of the drive takes place on I-35 which serves as the main hub of transportation in Austin.