Posted by Mustafa Hasanbulli - 26 June 2018
Auckland traffic has always been a hot topic in NZ. There are regular news articles about the woes of Auckland drivers as they face increasingly congested motorways, and it’s always a popular election issue.
Have you ever wondered what the traffic would be like if you leave now, or in another 30 minutes? For those of us driving to work, misjudging traffic flow and journey times can mean we’re late for that important meeting or miss those early bird car park deals.
Most of the time we can guess the traffic by previous experience e.g. if you are travelling to the city centre, it is fair enough to assume that it would be jammed in the morning and in the afternoon, as the majority of us start work between 7 and 9am and leave between 4 and 6pm.
This blog post is not about how bad the traffic is, or how we can solve it. It’s a fun experiment we set up out of curiosity to see if there was anything behind our assumptions. This isn’t a full blown modelling/simulation of Auckland traffic. If anything, we hope it shows how easy it can be to extract simple insights from publicly available data sources, using those insights to help with everyday decisions.
What we did
Google Directions API is a great tool to get estimates on how long it’s expected to take to reach a destination. We set up six Google Sheets spreadsheets with a simple script to make calls every minute to the API to retrieve travel duration estimates to/from three cardinal directions (north, west and south) to/from the CBD. As expected, there was a noticeable morning and evening rush.
Observations from Auckland commuter traffic
- The southern motorway seems to have very stable morning and evening traffic
- People living in the south seem to travel earlier than everyone else
- The northern motorway is the most chaotic in the morning (The Auckland Harbour Bridge will contribute to those higher travel times) and quietest in the evening
- It is worse to travel to the city between 7am and 8.30am. It is also worse to leave the city centre between 4pm and 6pm.
More interesting observations can be made if we break down the day-to-day travel times for each direction. Below are the plots representing weekday travel duration estimates:
From the plots above, we can make some observations:
- In most cases, getting to work on Mondays and Fridays has a similar pattern. We could assume the later Monday start could be because it’s more challenging to get out of bed the first day after the weekend. People may start later on Fridays or have the day off or work from home.
- We stay later at work at the beginning of the week and leave work earlier closer to the end of the week
All three directions exhibit the same behavioural trend: we seem to be really eager to leave earlier and earlier as the week goes on.
As we leave work earlier, travel times increase as well. So, leaving early might not be the best option.
If we look back at the overall travel duration times for three directions, we notice something else. Regardless of the direction you are going, if you are getting into the city in the morning or you are leaving the city in the afternoon, every half an hour you are late, you add a further 10 minutes to your total travel time.
This might sound like we are suggesting to leave early. Actually, we are saying the opposite. Leave later because travel times decrease faster after peak hours.
There is a linear relationship between the time you want to leave and how busy the traffic will be at the time
e.g. If you normally leave your house at 7am and you are planning to leave at 7.30am, this will add another 10 minutes to your total travel time. So, every half an hour is equivalent to 10 minutes additional travel time when getting into the city in the morning hours, and leaving the city in the afternoon.
Tips for Auckland commuters based on this data
- Don’t travel at peak hours if you can avoid it. These are: 6.30am - 8.30am if you are getting into the city, and 4.30pm - 5.30pm if you are leaving the city.
- We are causing the traffic jams on the southern highway by leaving way too early on Fridays, which defeats the purpose of leaving early
- Getting in at 6am and leaving at 3pm will give you nine working hours with the shortest travel times if you are travelling into the city in the morning and out in the afternoon
If you have any questions or would like to discuss how data science could solve problems and provide insights for your business, then please get in touch.
Mustafa Hasanbulli is a data scientist at Sandfield. He holds two doctorates in mathematics and theoretical chemistry. Since his career shift from being an academic to working in the business world, he aims to bring a different viewpoint into everyday business life using his theoretical knowledge and skills.
When not working, Mustafa enjoys quiet nights in with friends sharing good food and wine.
Make the most of your data
Today business decisions need to be made fast and often. With increased disk space, cloud services and internet speed a massive amount of data is now being collected in most industry sectors. This presents the opportunity for predictive analytics, data insights and data visualisation to assist in making data-driven decisions. At Sandfield we have qualified experts and tools ready to assist you with analysing your data.