The Problem With Traffic Infrastructure
Why I Started Strada Labs
There are more ways to get around a city than ever before. Scooters, bikes, rideshare, skateboards and even autonomous vehicles. New methods of transportation are popping up, often unannounced. The way we move is changing, and it’s getting harder for those designing our city streets to keep up.
The fact is that for the past century most US cities have been designed with the car as its default transportation option.
- 3 trillion miles are driven in the USA every year
- There are 272 million cars on the road
- 55 % of car trips are under 10 miles
This has resulted in a transportation infrastructure that has disappointed pedestrians, cyclists, public transit users and even the people in the very cars that cities were designed to accommodate.
People have taken note, and some are actively seeking alternatives to get around in the current transportation infrastructure. Lyft, Uber, Skip, Jump, Bird, Scoot, Lime — an ever growing number of companies are trying to improve the transportation experience while existing in an ecosystem that was not designed for anything other than a single occupancy automobile. In this process, pedestrians and cyclists often fall to the very bottom of the priority list. While crosswalks and bike lanes exist, they are often not respected or obstructed.
Movements like Vision Zero and Complete Streets are seeking to make our transportation infrastructure safe and equitable. Movements like these advocate for redesigning traffic infrastructure away from its current car-centric state and instead reward short trips, including walking, bicycle trips and use of public transit. Shorter trips often results in more vibrant cities, allowing for more social and economic exchange. Some cities are showing that they do have the will to change. San Francisco, for example, recently committed to another 20 miles of bike lanes.
But even a will to change for the better can result in unintended consequences. San Jose introduced a form of protected bike lanes that uses parked cars to separate automobile traffic from cyclists. The new parking areas and bike lanes were poorly marked and resulted in drivers parking in the bike lane, making it useless to cyclists.
Traffic Studies and Data Collection
Understanding traffic patterns is crucial in the creation of a better city. There are a variety of tools and techniques to better understand traffic. Traffic studies are diverse and vary greatly in scope, duration and cost. However, traffic studies have the same basic components which include talking to stakeholders, data collection, and suggesting changes or alternatives existing plans. The data collection portion of the study relies on vehicle and pedestrian counts and often includes more detailed data like turning paths. The data is collected over a predetermined period of time which is usually a few hours a day for a couple of weeks and is then extrapolated to represent a much longer period of time.
There are other ways of understanding traffic that do not require a full traffic study but most still employ some form of data collection. Data collection techniques also vary greatly. One of the most common and cheapest methods for data collection if to conduct manual counts. Assigning a person to an intersection or corridor to count pedestrians and cyclists is one of the most common methods of collecting data. Manual counts are especially useful when counting pedestrians, cyclists and other active transportation methods since these are not picked up by more traditional sensors. These traditional sensors include inductive loops or pneumatic tubes which are laid on the road and sense when an object travels over it. These types of sensors have a hard time differentiating between types of vehicles but can be deployed for much longer than a human counter.
“Sometimes we don’t even count those”
— PE Ian Lockwood, referring to walking trips
There are many other ways of counting traffic that are better suited for active methods of transportation. Some of these methods include tracking MAC addresses of smartphones, buying data from activity tracking apps, and using cameras with object detection software. Tracking MAC address, also known as Wi-Fi sniffing, relies on the deployment of small devices that are able to detect wi-fi signals, using smartphones as a proxy for humans. While this is a clever way of counting pedestrian traffic it relies on the assumption that every person has a smartphone. This is not always the case. Many people do not have smartphones and on the other hand people increasingly carry multiple devices that can be “sniffed” such as smart watches or multiple phones.
Getting data from activity tracking apps such as Strava, is particularly clever because it does not require any sensor deployment or infrastructure. Users upload detailed data about their activities which include their path, speed and transportation method when they are out for a run or riding a bicycle. However this data is not representative of the population because it focuses exercise or recreational trips which are a very small subset of total active transportation trips.
Using cameras for data collection is becoming increasingly popular, but it’s not necessarily new. Many DOTs have been using cameras to monitor traffic for many years. They’ve also been used on tolls to read license plates and for automated enforcement of red lights. Recent advancements in computer vision, artificial intelligence and computing hardware are enabling the use of cameras for much more detailed data collection. Computer vision software is now able to detect almost anything a human can, making it a great option for multi-modal data collection.
They biggest drawbacks for camera based data-collection methods are costs and privacy concerns. Many of the companies offering camera based solutions require their own hardware and mounting infrastructure to be installed which add considerable cost to their deployment. This technology is also vulnerable to a number of attacks and abuses, which has led to some of its use cases, such as facial recognition, to be banned or regulated in a number of cities. Despite its drawback, camera based methods offer the most detailed and robust traffic data.
An overview of active transportation data collection systems by Alta Planning + Design can be found here.
Strada Labs
Data collection is a very important component of urban planning and transportation management decision making. Planners and engineers make decisions and recommendations taking into account the data that is available to them. There are many methods and tools to acquire traffic data but most of them have significant drawbacks which leave much to be desired. I believe that with the right tools and methodology the datasets available to planners and engineers can be greatly improved, which would result in a more equitable city infrastructure. I started Strada Labs to fill the gaps in the existing data and help cities, planners and engineers make better and more informed decisions that will lead to a positive change to our city infrastructure.
I’ll publish a more detailed post about what we are doing in the next couple of weeks, but for now check out our website www.stradalabs.com and stay in touch!
At Strada Labs we are working toward providing urban traffic data that is accurate, continuous, multi-modal and real time. Our camera based approach to data collection leverages existing transit cameras, greatly reducing cost. We also provide our own camera and processing hardware for locations in which there is no existing cameras. Our multi-modal solution includes anonymized counts, paths taken and speed estimates for pedestrians, cyclists, buses and cars. We also provide a dashboard to help better visualize our data and help detect patterns and trends.
I’m planning a future post that goes more in depth into our work at Strada Labs, but for now check out our website www.stradalabs.com and stay in touch!
