How smart IoT will disrupt how regulations are enforced
Civic analytics—harnessing data analytics capabilities to improve government services—is not wholly new. For example, smart tax systems are improving tax compliance and collections. Analytics are helping to police financial crimes, such as money laundering and tax fraud. Other government systems use civic analytics for welfare and employment, public security, transportation, and economic development.
But those are just the first steps. "Government data initiatives are fueling a movement toward evidence-based policymaking," according to a McKinsey report. "Data enablement gives governments the tool they need to be more efficient, effective, and transparent while enabling a significant change in public-policy performance management across the entire spectrum of government activities."
Among the changes you should expect in public-policy performance management are AI-assisted, layered compliance enforcement. For example, according to the report, smart traffic systems that control traffic lights and routing to relieve congestion could also do the following:
- Issue tickets to drivers who are not in compliance with traffic and air pollution laws.
- Render vehicles inoperable for drivers who have not paid prior traffic and parking tickets.
- Inform road maintenance crews of potholes and predict future road damage. In addition to making repairs, the system could fine vendors for not meeting requirements.
- Recommend regulation modifications to lawmakers based on driver, vehicle, and road performance data.
Eventually, AI will use Internet of Things (IoT) and other data and analytics to not only enforce compliance and modify regulations to optimize results, but actually write regulations as well.
"Governments need to launch data initiatives focused on, among other things, evaluating policy performance, reconciling quantitative and qualitative data, and allowing the implementation of a continuous improvement approach to policymaking and execution," according to the researchers.
Google’s AI already can create better machine learning code than human developers. If a machine can already determine the code it needs to do its own job better, it’s not far-fetched that AI can write regulations to improve human outcomes and conditions.
Although some argue the extent of AI’s ability to create policy is limited, multiple cities are already using data and analytics in their regulatory regimes to enable increased efficiencies and lower costs for taxpayers. Incumbent protection is washing away in its wake, and so too are a number of traditional partisan approaches to policymaking as data takes center stage in decision-making.
It appears to be only a matter of time before AI-based automation assumes the work and takes over much of the formation of practical regulations, as well as tweaks and adjustments to existing regulations as changes in the data dictate. That’s easier to imagine if you consider how cities are using data and analytics now and how easily those processes can be assumed by AI.
"Just last month, New York City announced a reduction in fines imposed on small businesses because too few citations actually had anything to do with health and safety," reports Stephen Goldsmith in a Harvard article. "And Chicago's path-breaking project that analyzes data to identify bad actors and concentrate on them—allowing food inspectors, for example, to prioritize on establishments in need of more scrutiny—promises better service to businesses and consumers alike."
A Catalog of Civic Data Use Cases, published by Harvard University, provides a list of city operations enhancements that are currently data-driven. The four main categories of operations are health and human services, infrastructure, public safety, and regulation enforcement. Analytics are currently heavily deployed to aid with city services, in and between these civic operations, in almost all cities whether or not they earn the distinction of a "smart city."
Since cities also routinely compete as a group, in regions, connectivity via shared data and joined analytics between cities and throughout states also likely will follow. The spread will continue throughout the country.
This does not necessarily mean that human representatives in government will be replaced by AI. Indeed, it’s a matter of considerable debate as to where and when to draw the line in AI’s jurisdiction over human affairs.
Even so, McKinsey predicts that smart cities will completely redesign how states (and by extension, the country) operate. That’s not likely to happen in the near term.
Instead, governments will use civic analytics and smart city data to accomplish more with less.
"Although authorities may be wary of giving up control," says a second McKinsey report, "great city leaders accept that not every service, from information services to park maintenance, needs to be provided directly by government personnel, and they acknowledge that the mandate for government changes over time."
Learning on the fly
Like most evolutions, the change in city dynamics isn’t very orderly. But as with so many technologies that earn the label "disruptive," smart city evolution is happening very fast.
In short, both humans and machines are learning on the fly.
Take drones as a current case study. U.S. cities are actively working to implement large-scale drone programs to support many municipal functions, said Brian Russell, co-author of the report "Establishing a Safe and Secure Municipal Drone Program" and chair of the Cloud Security Alliance’s IoT Working Group, in a statement. Municipal drone uses range from medical, transportation, and agricultural to emergency management and infrastructure protection.
However, civic drones aren’t the only drones buzzing about in smart cities. Private sector drones are in the air as well, both commercial (as with Amazon Prime Air and pizza delivery) and consumer class (everything from adult hobbyists’ machines to little Emily’s Christmas present). It’s the latter that can complicate civil services, as happened recently with potentially fatal interference with firefighters during the California wildfires. A smart city drone network has to incorporate sub-networks such as commercial deliveries, city and national security, emergency response, and infrastructure maintenance.
"Drones in the sky, drones in the sea, drones on land. But are we ready? The mass adoption of drones by cities implies that thousands of programmable connected mobile devices will not only operate in the streets, but also above and below them," warns Russell.
Regulation and enforcement are necessities for all of these drones to coexist and operate safely, and they need to be flexible enough to respond to this fluid environment. Fortunately, data collected in real time from drones and other smart city devices can both inform public policy and help enforce public policy.
"Any use case—even outside of a drone program—can help shape regulation. The connected landscape is changing rapidly, so everything is still shiny and new. The only way to shape regulation is to learn from our failures and successes," says J.R. Santos, executive vice president of research at the Cloud Security Alliance.
Data vs. turf, knowledge vs. will
In theory, continuous learning with data gathered from drones, autonomous cars, infrastructure robotics, and other smart city IoT data should be a straightforward exercise. Further, the knowledge collectively gleaned from each of these sources provides greater insights into efficient means to regulate their use as standalones and as elements in a larger system.
The debate on whether this should be allowed to happen includes a growing concern over the burdens that frequently changing regulations may place on businesses and citizens. Advocates say AI will automatically adjust operations on the civilian side to comply with evolving regulations, and thus any burden would be slight and worth it compared with the potential benefits. Critics say AI lacks human intuition, and thus its judgements may be impractical and burdensome in practice.
In any case, enforcing regulations should be data-based and thus equally applied. After all, sophisticated algorithms can automatically determine compliance easily enough. But there are significant differences in how the new versus the old algorithms function.
"For example, if smart building and waste management systems were implemented, a rigid approach would be to define the regulations up front and then hard code them into the systems," explains Rohit Talwar, CEO of publisher Fast Future.
"By contrast, a learning approach would use the accumulating data on usage patterns, waste levels, energy consumption, air quality, and emissions to determine whether any critical limits were being breached and also to identify best achievable practice. The regulations could then be adjusted over time to drive users towards best practice," Talwar adds.
Unfortunately, the reality is not nearly so logical and organized.
"People can learn from any use case, but since some devices may have different regulatory bodies, each class of device may have to be handled separately," says Santos. "For example, the FAA might be more interested in the drone program while the FHWA may be more interested in connected vehicles."
In other words, it may not be clear which agency has jurisdiction in every case. In either scenario, turf protection could soon become a problem. In other cases, government agencies with insufficient budgets and manpower may elect to push the problems on another agency’s to-do list.
In any case, lawmakers have a poor track record in keeping up with technology and may be more interested in protecting the status quo than in leveraging smart cities. Then there is the age-old problem of lobbying and pushbacks from special interests that can get in the way.
"Data is data, ultimately. How it’s interpreted, on what pace, and whose ox is gored as a result is another matter—and those groups who feel aggrieved will react as quickly as any corresponding legislative or agency initiative," says Richard Santalesa, attorney at the SmartEdge Law Group.
"And if the action is on an agency level, there’s always the generally required rounds of public comment and update to enact regulations," Santalesa adds. "Add in that most of the IoT initiatives have potentially significant privacy implications and require the glacial fed/state government contracting process to grind their slow gears to launch such, and I’m not seeing any brave new smart city IoT world springing forth like Athena from the head of Zeus fully formed."
In the end, it is likely smaller government budgets and bulging populations will bring an end to the resistance to change. Voters are not likely to accept the alternative, which would be drastic cuts in government services. Like all disruptors, smart cities will change everything despite the objections of former giants and power brokers.
Smart regulatory compliance: Lessons for leaders
- Expect existing analytics for regulatory compliance to adopt machine learning and AI at an accelerated rate. This class of analytics simplifies evolving regulatory changes in governments all over the world and thus reduces the time it takes to comply. These analytics will take on even more importance as more smart cities come online and lawmakers seek to capture and leverage the benefits.
- Consider whether data from your company’s IoT use can be further monetized by allowing cities to use it in their analysis of the need for new or modified regulations.
- Proactively work to ensure smart city data is and remains open for commercial use. This data can help inform your company’s operations and train your own AI, and predict changes in the evolving regulatory environment.
This article/content was written by the individual writer identified and does not necessarily reflect the view of Hewlett Packard Enterprise Company.