4 min read

Analyzing the water P3 market, how AI is already powering transportation planning, and debunking AV parking myths

Analyzing the water P3 market, how AI is already powering transportation planning, and debunking AV parking myths

A recent report from EY and the American Water Works Association exclusively focuses on P3s in the municipal water market sector. The report makes the case that the need for advanced water infrastructure in the US will only increase in the years ahead, and granting P3 concessions are a good solution for municipalities in many circumstances. (AWWA is a non-profit with 52K members, including over 4200 utilities that supply roughly 80% of US drinking water and treat almost 50% of domestic wastewater).

As the introduction to the report notes, "relatively little has been heard from those practitioners directly involved in planning and running municipal utility systems. . . . [the report] therefore explores the perceived attractiveness of P3 as an alternative delivery model from the perspective of those directly responsible for the nation’s water infrastructure."

Water P3s in the US have faced headwinds over the last decade - the last to reach financial close was the San Antonio water supply pipeline P3 in November of 2016. (There have been several water system concessions - the cities of Rialto, CA and Bayonne, NJ, for example - but the report is focused on the development of individual water assets rather than entire systems, although these of course can derive benefits from P3 models too.)

The slim track record of successful water P3s has left stakeholders leery about the pros and cons of P3s. Water assets in particular also face heightened political scrutiny over a perceived loss of municipal control over the underlying asset. A lack of P3 champions therefore exists among owners of water assets who understand the financial, legal, and procurement challenges to launching a successful P3.

But the report does suggest that attitudes are changing. In part this is thanks to an overall shift in the market to design-build and more progressive procurement methods generally, coupled with increased federal funding through programs like WIFIA (the water equivalent of the TIFIA program in the transportation sector).

Municipalities believe that water treatment infrastructure, advanced water reuse/recycling and bioenergy/biosolids treatment are the most suitable assets for a P3, given the higher amounts of capital, technical capacity, and operating risk that is inherent in these types of projects. P3s can also generate innovation for these highly technical projects as well as create value-for-money for the owner through the construction and performance risk transfer that is inherent in a P3.

There is little doubt that the US water sector will need to invest over the coming decades in increasingly complex infrastructure. And P3s will undoubtedly have a larger role to play. But as the study also points out, "unlike other P3 markets, such as Canada and the UK — there is a relative lack of objective guidance that municipal decisionmakers can utilize to inform their thinking and to build stakeholder support for considering P3." Water industry stakeholders will need to step up to the plate and, as the study concludes, provide "further guidance from trusted and independent sources" about P3 best practices to help municipalities benefit from the model when it makes sense. (EY, AWWA)

AI chatbots are the rage right now. So it's easy to forget that AI is already being used by civil engineers and urban planners to improve transportation and transit systems beyond, for example, the promises of autonomous vehicles, eVOTLs, and other Jetson-like, futuristic technologies.

Connected technologies, intelligent transportation systems, and better collection methods (GPS hardware, remote sensors, etc.) are all providing engineers with better data to understand how people and goods move through the built environment. More precise data means our transportation network can be less "general" and used to target both greenfield projects or improvements at a granular level for specific populations or "nodes" within a given network. (Think about the asynchronous commuting patterns in many urban areas post-pandemic, for example.)

AI is also extremely powerful when applied to supply chain logistics. It can inform route planning and freight vehicle departure times based on historic traffic patterns, weather data, and other general, temporal, or location-specific inputs that, when optimized, provide a better network user experience for drivers, customers, and other users active on the network. This type of analysis is ideal for AI tools, which can learn what works and what doesn't without an engineer building a complex model that must be tweaked over time.

But the real power of AI as applied specifically to transportation systems is in how it allows engineers to move beyond designing based on "aggregated" data - like monthly or annual traffic volumes and population changes - to much deeper levels of detail.

Over time, and assuming it doesn't conquer us all, AI should help the industry better develop and serve the US transportation system at large, and for specific regions and populations. And, as data collection and processing methods continue their advance, more powerful AI models will provide even better information to determine the best solutions for a given community, system, or traveling cohort. (ASCE)

The potential for autonomous vehicles to reduce the number of parking garages, lots, and spaces in urban and suburban locations might be exaggerated. A recent study from the Urbanism Next Center at the University of Oregon, funded by autonomous vehicle company Waymo, recently reached this somewhat surprising conclusion.

The argument that AVs could eventually free up dramatic amounts of urban and suburban land for redevelopment has historically been based on theoretical models of hypothetical cities (in some cases predicting up to a 90 percent reduction in parking demand). Instead, the Waymo study's researchers modelled different scenarios in actual San Francisco neighborhoods based on AVs replacing cars. They asked whether the drop in demand for parking would, in turn, cause the reclaimed spaces to be redeveloped into housing or other more productive uses than garages, on-street parking, or surface lots.  

In already dense, expensive neighborhoods there is already sufficiently brisk demand for new housing. So parking doesn't really enter the analysis; it's instead a political challenge that cities must solve in order to build more housing. But in the suburbs or car-centric cities (like LA, Dallas, or Phoenix) with spotty mass transit the study concluded that a 40% drop in parking demand could yield up to 3680 new housing units, and a 60% drop up to 5600+ units. Achieving these reductions would still require all AVs to be shared rather than owned (which 54% of Americans say they won't do) and all AV rides to be carpools (which is also unlikely).

These mixed results, which the study acknowledges will depend on density and geographic location within a particular region, suggest that the real estate development impacts of AVs are still very much an open question. They also highlight that cities should carefully consider the costs of deploying a region-wide AV network before redirecting them from other transit priorities. (Streetsblog)