The concept of Smart Cities offers the promise of urban hubs leveraging connected technologies to become increasingly prosperous, safe, healthy, resilient, and clean. What may not be obvious in achieving these objectives is that many already-existing utility assets can serve as the foundation for a Smart City transition. The following is a broad discussion on the areas of overlap between utilities and smart cities, highlighting working knowledge from experience at PG&E.
Smart Cities and the Weather
The views set out in this article are my own and do not necessarily represent the positions, strategies or opinions of my employer, IBM.
A study by the US National Center for Atmospheric Research (NCAR) in 2008 found that the impact of routine weather events on the US economy equates annually to about 3.4% of the country’s GDP (about $485 billion). This excludes the impact of extreme weather events that cause damage and disruption – after all, even “ordinary” weather affects supply of and demand for many items, and the propensity of businesses and consumers to buy them. NCAR found that mining and agriculture are particularly sensitive to weather influences, with utilities and retail not far behind.
A moment’s thought will confirm that many city systems are likewise routinely influenced by the weather – systems ranging from energy and water, to sanitation, transportation, healthcare, parks and recreation, policing and so on. Add in extreme weather events, and the list grows to include event/disaster forecasting, preparation and response.
Many of these, disaster management included, are the focus of smart city innovations. Not surprisingly, therefore, as they seek to improve and optimize these systems, smart cities are beginning to understand the connection between weather and many of their goals. A number of vendors (for example, IBM, Schneider Electric, and others) now offer weather data-driven services focused specifically on smart city interests. For example:
- In the energy field, prediction of demand, renewable energy yield, and storm-related outages, to enable grid configuration and management, including demand-response management;
- With water systems, prediction of demand, management of irrigation activity and control of such tasks as water aging and blending, chemical dosing, and waste water treatment;
- In transportation, adaptive traffic controls and routing, drive time estimates, fleet disposition and management, and road safety;
- In environmental management, prediction of air pollution (ozone, particulates, NoX) and water pollution;
- With smart buildings, inputs for adaptive HVAC and free-air cooling;
- Prediction and management of disruptions, for example from rainfall or snow – and at the other end of the range, forecasting severe weather events that lead to risk and damage.
The underlying trend here mirrors what is happening in cities themselves. First, as with smart cities, it’s about the Internet of Things and the impact that it has on weather forecasting capabilities. Satellite data is one part of this, but terrestrial data is a huge component also. IBM’s Weather Company alone uses over 250,000 weather stations globally, many of them from its citizen-sensing network, the Weather Underground; and it processes over 100 terabytes of 3rd party data per day. Southwest Airlines’ planes now detect atmospheric water vapor levels for NOAA. The Con-Way haulage company collects weather data on its trucks as part of the US MesoNet program. I cannot prove it, but it seems reasonable to suppose that collectively, weather data collection may be one of the largest uses of the IOT to date, in terms of both the number of collection points and the volume of data.
Second, and as with smart cities, it’s about the ever-growing power of analytics and AI. Weather models have long been among the major uses of super-computing resources. However, the ability today to create high resolution or micro-forecasts (at a scale as focused as 0.2 miles/0.5km) effectively takes weather forecasting to a whole new level of capability and application, especially when combined with other data on topology, traffic, vegetation and so on. In my previous blog post, I mentioned just such an example of this with the AI-powered air pollution forecasting system in Beijing that differentiates pollution levels on a micro-scale; a similar level of precision micro-forecasting, when combined with other data, makes it possible not just to forecast output from wind-farm or solar installation many hours in advance, but for larger installations output by zone within the installation, at >90% levels of accuracy.
And yet, while utilities and other “super users” of weather data are certainly beginning to understand the precision and value of highly focused, customized micro-forecasts, one has the impression that cities still tend to think of “the weather” as something that applies to the whole city in a one-size-fits-all manner, and where forecasts are largely consumed from the TV or a computer app. Micro-forecasting (perhaps purchased from the local energy utility, who may already be using it!) could therefore be an astounding resource. Further to the examples above, some possible use cases (which do not yet exist, to my knowledge) for micro-forecasting might include the following:
- Probabilistic snow forecasting could be linked to on-the-ground data that allows the city to assess where best to station its snowplows, taking account of the forecast pattern of snowfall, the location of critical assets such as hospitals and traffic hotspots, and the maintenance status (and thus availability) of snowplows at that time. It could also be linked to a work order dispatch system for the crews.
- Hyper-local temperature and humidity forecasting/observations could allow a city to predict the energy load from its own buildings (and perhaps others in eco-districts or public housing) given their known performance, and thus the margin available for demand response purposes. The forecast could also be used to determine where extra cooling may be required for senior citizens or disabled residents. This system could then drive the necessary grid configurations and perhaps neighborhood support for the vulnerable.
- A fog micro-forecast might be linked to accident data to predict the likelihood of traffic accidents by location, and thus drive the configuration of variable speed limit signs, and perhaps (as is already being piloted) the hue of smart street lights in order to reduce glare.
- A temperature micro-forecast could be linked to a water system’s SCADA system to pre-plan pumping activity to maintain quality and water mixing at least energy cost. As local or micro-forecasts of evapotranspiration (ET) are increasingly used to drive irrigation activity, these could also be fed in to help predict demand by district meter or pressure zone.
- A wind and temperature micro-forecast could, when combined with topology and vegetation data, allow first responders to plan where best (and most safely) to position firefighters to deal with a wildfire.
One’s imagination could run on in this vein for some while! The key to deriving value from micro-forecasting, however, is to ensure that the forecast is integrated in as automated manner as possible, with the business process that executes the response. “Rip and read” interfaces risk adding delay and human error that undermines the value in the accuracy and precision of the micro-forecast itself. The forecast needs as a minimum to be a direct input to control room systems, and to deriving recommended actions through decision support tools, even if the actual response is human-actuated. One suspects that this may be a little way off in cities, while city staff learn to trust the micro-forecasts – but industries such as utilities and airlines are in many cases already there.
Leave your comment below, or reply to others.
Read more from the Meeting of the Minds Blog
Spotlighting innovations in urban sustainability and connected technology
When the idea of smart cities was born, some ten to fifteen years ago, engineers, including me, saw it primarily as a control system problem with the goal of improving efficiency, specifically the sustainability of the city. Indeed, the source of much of the early technology was the process industry, which was a pioneer in applying intelligent control to chemical plants, oil refineries, and power stations. Such plants superficially resemble cities: spatial scales from meters to kilometers, temporal scales from seconds to days, similar scales of energy and material inputs, and thousands of sensing and control points.
So it seemed quite natural to extend such sophisticated control systems to the management of cities. The ability to collect vast amounts of data – even in those pre-smart phone days – about what goes on in cities and to apply analytics to past, present, and future states of the city seemed to offer significant opportunities for improving efficiency and resilience. Moreover, unlike tightly-integrated process plants, cities seemed to decompose naturally into relatively independent sub-systems: transportation, building management, water supply, electricity supply, waste management, and so forth. Smart meters for electricity, gas, and water were being installed. GPS devices were being imbedded in vehicles and mobile telephones. Building controls were gaining intelligence. Cities were a major source for Big Data. With all this information available, what could go wrong?
If you want a healthier community, you don’t just treat illness. You prevent it. And you don’t prevent it by telling people to quit smoking, eat right and exercise. You help them find jobs and places to live and engaging schools so they can pass all that good on, so they can build solid futures and healthy neighborhoods and communities filled with hope.