If Canada’s urban centres want to become leaders in digital transformation, they need to agree on a set of data standards, according to a leading University of Toronto engineer, while making a concerted effort to identify both new and existing challenges along the way.

Giving the keynote address at Technicity, a Dec. 7 event co-sponsored by IT World Canada, U of T urban systems engineering professor Mark Fox discussed with both tech industry veterans and City of Toronto staff how they could best use big data and other types of analytics to digitally transform their city.

“I attend meetings, give talks, et cetera, around the world, and I always hear various people talk about the virtues of the digital transformation of cities,” Fox told the audience. “But they tend to provide a very rosy picture of where we are heading, and tend to gloss over some of the difficult pieces of what that transformation entails.”

“Cities are not rocket science,” he told the crowd. “They are a thousand times more complicated than rocket science.”

In science, he said, the goal is to reduce complexity through design, while cities contain so many elements – education, public safety, water and sewage, recreation – that although each needs to be considered, it’s impossible to fully incorporate them all.

During his presentation Fox called the challenges facing digital transformation “potholes” and “sinkholes”: the former, he said, are problems that can be fixed easily enough, while the latter require completely rethinking how the infrastructure is created in the first place.

Asked for examples, Fox told ITBusiness.ca that data sparseness – recognition that the data collected is only a small part of what’s needed – is a pothole, while data standards – “coming up with attributes and values that are inter-operable not only within a city, but across cities” – represent a sinkhole.

In theory, planners know the causes behind data sparseness and how to correct them – for example, the existing data collected by many City of Toronto departments often remains with whatever department conducted the research in the first place and becomes inaccessible to everyone else, the result not of privacy concerns but lack of planning.

“The reality is the data that we collect is only a very small subset of what we need, and I don’t know to what extent when we start anything we really ask the question: What do we need?” Fox said during the presentation.

A similar lack of planning led Ontario mass transit authority Metrolinx to neglect the opportunity to use its recently-installed Presto machines to collect exit data, he noted.

The problems caused by these potholes can often go back years, he said – for example, Toronto first announced that it would begin collecting traffic sensor data after installing the first automated traffic lights in 1967 – and has made none of it readily available since.

But even the oldest pothole pales in comparison to the inherent problems behind what Fox called called “behavioural characteristics” – the systems that city planners build to actually collect and deliver data.

“Digital transformation is difficult,” he told ITBusiness.ca. “There are a number of problems that we have to overcome… beginning with the data itself, and ending with our overall expectations of the behaviours those systems are able to display – and I’m really concerned with those behavioural characteristics in the long run.”

For example, during his presentation Fox showed the audience three official maps of Toronto that government officials use for data collection purposes – the city’s neighbourhood map, followed by the Toronto Police Service’s district map, followed by the federal government’s census division map – none of which shared any boundaries, and therefore measurable data.

Now imagine trying to share the data collected between multiple cities, Fox said, emphasizing again that inter-operability is impossible without data standards.

And so, for the second half of his presentation Fox laid out four behavioural guidelines that researchers should keep in mind when establishing standards for digital transformation.

Guideline #1: Awareness

Simply collecting and releasing data is not a worthy goal in and of itself, Fox said: there needs to be a reason for choosing which data to collect and release. Making those choices requires awareness – knowing what to expect from a given data point, its limits, whether deviations are significant, and which actions to take when deviations occur.

Guideline #2: Responsiveness

Digital transformation standards require responsiveness, Fox said: Being able to flexibly respond to events with a focus on outcomes and not the methods behind them.

In particular, he said, programmers and researchers should ask themselves the extent to which a given response incorporates:

  1. Situational Understanding, ensuring all relevant information is available;
  2. Shared Knowledge – that is, access to prior experiences;
  3. Flexibility, incorporating versatile responses to multiple situations;
  4. Empowerment, both for the system and the people behind it;
  5. Teleology, by replacing a set process with tangible goals.

Guideline #3: Introspection

This guideline encourages digital transformers to examine their new system’s performance and identify new ways to achieve its goals, by asking if the system can:

  1. Recognize Failure – that goals are not being achieved;
  2. Diagnose the root cause of the system’s current performance;
  3. Construct better solutions;
  4. Predict future events or behaviours;
  5. Be considered Self-aware.

Guideline #4: Accountability

Cities must be held accountable for their actions, Fox said – which means their systems must be as well. That means digital transformers must be able to:

  1. Determine Responsibility by ensuring the system leaves a digital trail of information they can follow to identify the cause of a particular decision or action;
  2. Determine the Underlying Basis for decisions by identifying and analyzing recurring patterns of behaviour;
  3. Appoint a Digital Ombud – someone the system is accountable to.

Conclusion

When it comes to potholes, Fox acknowledged there is no magic bullet.

“It’s hard work,” he said. “I think the first step is awareness that these problems exist, and… identifying the areas you have to prioritize if you want to get rid of the sparseness in your data.”

“Cities don’t have infinite funds, or infinite people,” he continued. “So the question becomes identifying the areas in which we want to have a more complete understanding, and the data necessary to develop that complete understanding.”

As for sinkholes, Fox said he hopes that future software vendors hired by Toronto – or any city – meet his four guidelines before either party signs a contract.

“The city’s not going to build its own systems all the time – the vast majority comes from enterprise software vendors,” he told ITBusiness.ca. “To what extent can we expect accountability? To what extent do we have digital traces of what goes on within the system? To what extent can we use those traces to actually track down the root causes of poor execution, or decision-making?”

“These are the elements that we need to build into our cities’ enterprise software systems,” he said.

Share on LinkedIn Share with Google+
More Articles