Licence to GovHack
Patrick Drake-Brockman took the plunge and went from data analyst to hackathon contestant. He shares his GovHack experience working on a solution to help learner drivers get their licence more easily, using open data.
Last year, I decided to test my skills in a hackathon. I signed up as an individual participant, a GovHack ‘single’ as it were, and had no idea of what to expect. By the end of the weekend, I’d made new friends, tested my abilities and my team won an award.
For a newbie like me, the Connections Event was invaluable. It explained how GovHack runs, where I needed to be and when and how to submit our entries. The organisers brought together a cross-section of mentors and volunteers to help us get ready and guide us on the day.
More importantly, singles like me found a home within a team of other GovHackers.
Fueled by coffee, my team gathered on Friday night to plan our hack. In the ACT the government is building a light rail network, so the Canberra event had a strong focus on transport. But we didn’t want to build another bus tracker or bike path app.
During our discussions, one of the younger team members remembered how costly it was to get a driver’s licence. We started to think how an app could help a learner driver build their experience and competence, help their supervisor teach them, and perhaps even save some money for those learning to drive.
As some of us are parents with kids fast approaching the learner driver age, it made us think how do we teach them? After years of driving, we do not consciously focus on driving competencies as these are now ingrained. With this in mind, we designed the app with the supervising licenced driver in mind, as drivers cannot use a mobile phone. We needed an app that clearly sets out the driving competencies so that both the learner and their supervisor has a shared understanding of what is being practised. We also wanted an app that locates safe places to drive and pinpoints the locations where these driving competencies will be tested. We called the app Licence to Learn, and it has elements that support the learner, including a revision of driving competencies before you get in the car and map directions that can be spoken while driving.
To define the data for this app, we evaluated what the user needed to know. Learner drivers must demonstrate specific competencies to earn a licence, so knowing what those were was key. We also knew that learner drivers can feel nervous and unsure, as their road knowledge is developing.
Our team investigated safe roads information and looked at data on pedestrian, car and cyclist crashes. We felt this could give our users more information ahead of time of particularly tricky intersections or ‘black spots’.
We also ‘pinned’ driving competencies to a map by geolocating their associated road signs using the ACT government’s road signs dataset.
Our team used road use density information in our app as an important way of identifying safe areas for a beginner, intermediate and advanced learner. Unfortunately, we couldn’t locate an open traffic density dataset. We felt this information was important as it would help our users increase their confidence and allow them to avoid heavy traffic. However, we were able to infer this information using the ACT Government’s mobile speed camera data. We built an algorithm that estimated road density by treating mobile speed cameras as a random traffic sampler, averaging vehicles numbers checked per time period to approximate traffic density.
Finally, we brought together the open data with the driving competencies in an app that we aimed to make simple and easy to use.
For mapping, we relied on the inbuilt Google travelling salesman approximations (to find the shortest route) rather than build our own. This sped up development and let us leverage Google’s infrastructure to plot driving routes.
We didn’t expect to win the main transport prize, but we were pleasantly surprised. With GovHack coming up, we are “getting the band back together” to have a tilt at it once again.
You can view our project repository and video at Hackerspace.
The DTA is a lead agency sponsor of the 2018 GovHack.