Transport Data – the key to solving the challenges of today and planning for tomorrow

Data is king, and insights derived from it are gold. However, its value is significantly constrained if its collection and use are not planned. That was the key take away from the Data Insights and Approaches forum I recently chaired at ITS’ Mobility 2020 – MaaS, Future Mobility & More.

The webinar comprised of some of the best voices in the mobility data industry, featuring Professor Fang Chen of the University of Technology Sydney (UTS), Andy Taylor of Cubic Systems, Dr Hanna Grzybowsky from CSIRO and John Cardoso from Intelematics to talk about the potential benefits we are seeing coming out of the use of transport data, as well as some of the common pitfalls and limitations.

The first idea raised was that transport data could solve broader issues than the challenges relating to transport services. This idea was presented by Professor Chen, who’s recent work at the University of Sydney has been to look at the effect of COVID-19 clusters on public transport and vice versa.

Professor Fang’s work concluded that in areas with high clusters of COVID-19, there was a reduction in bus delays as these cluster areas saw less traffic and fewer people taking buses. Whereas in areas that had lesser COVID-19 clusters, traffic and bus use was higher, leading to more delays. The study also concluded that traffic activity remained higher in Western Sydney, as it housed more essential services compared to the East, where population density is higher and work from home more common.

Dr Grzybowsky from the CSIRO noted that COVID-19 has created challenges for research, putting the collection of data in a holding pattern as we learn what the new normal is. The movement of people has altered, but will it ever go back to what it was? Or would we have to navigate through more inconsistency? We still don’t know. Intelematics’ John Cardoso noted that with such disruption, the historical travel data becomes even more critical as it can be used as a ‘baseline’ to plan from and understand how people move.

Predictive tools can then try and map out potential modelling, which as Dr Grzybowsky mentioned, can at least help with contact tracing. Europe is currently leading in this space due to the head start they had in tackling COVID-19 earlier than Australia. But as Dr Grzybowsky noted, a framework needs to be set to strengthen public confidence.

Frameworks are important. As Andy Taylor from Cubic Systems pointed out, traffic data can be highly beneficial and real-time traffic avoidance systems – backed by data – could save up to 730 million lost hours per year. But Taylor noted the serious risks of getting it wrong, with privacy regulators around the world beefing up their resources and pouncing on companies that do not comply. An opportunity can easily turn into a liability – and quickly. 

 Taylor encourages businesses to know their risks: 

• Privacy implications – What are the laws? What have you promised your customers? 
• The validity of data – How valid is the data you are using for your purpose? 
• What are the reputational risks if you get privacy and validity wrong?  
• Do you have a plan? Who verifies it? Who is managing the process? 
•  Do you have an inventory of your data, so you know what you have in case something gets lost? 

Taylor noted that even companies as large as British Airways have failed to ensure they knew what they were getting into. 

Being realistic is also a point Taylor emphasises. Too often, companies ask for more data from individuals than they need. This creates scepticism around a company’s true motives, inviting regulators to probe, which creates unnecessary management costs.  

Capturing too much also runs the risk of data being fragmented across different platforms, with Taylor reminding us that a person has the right to be forgotten in many jurisdictions. If the data is sitting across multiple systems, it could become a logistical nightmare to manage, leading to compliance issues further down the track. 

Professor Fang also added that in her view, privacy is the biggest challenge to overcome in the widespread use of transport data. Therefore, it needs to be treated as a ‘risk’ for mindsets to change and confidence granted. 

Taylor encourages companies to understand global data frameworks as they differ in focus between regions. The EU – considered to have the most advance framework – bases its framework on three pillars: governance (public vs private interface); transparency (including consumer buy-in); and management (the operating models used). Meanwhile, the US is still in the process of building a coherent data strategy. In this sense, Taylor cautions governments not to design systems that give them the right to request data from a private entity on whatever they choose, to do so would run the risk of breaching commercial confidence and undermine trust. 

It’s no surprise that data is highly valuable. But how does its value work?

The value of the data varies according to the use cases. The currency – or how recent the data is – usually plays an important role when evaluating the value of the data. This is especially relevant for operational activities such as incident management, where you need to know that an incident happened and the location as quickly as possible so that you can intervene. However, when trying to understand if this incident was to be expected, or an outlier, you will need to look at historical trends. Therefore while the unit value of the data may decline throughout time for operational use cases, for automation/machine learning or simply for trend analysis, the data only has value if it has enough scale. Cardoso added that when thinking about the value of data, you should not only think of its immediate unit price but also on its scale and use cases.

The great thing about chairing forums like this one is how much you get to learn from experts sharing their experiences and unique points of view. We’re at an interesting point in time as the world wakes up to the real value of data, but at the same time, approaches data collection with healthy scepticism. It is paramount that we continue to have these discussions so we can adopt the best practices and realise the benefits of collecting and using data while mitigating risks involved.