Emerging technology changing the face of public transit post-pandemic

By KATE COIL
TML Communications Specialist

Emerging technology stands poised to change not only how city buses operate but how those buses could end up playing a role in reducing traffic congestion, parking enforcement, and more.


Government Technology and its parent company eRepublic recently held a webinar titled “How Smart Transit Tech Will Get Cities Moving Again” to discuss ways new technology will meet demands for public transportation spurred on by lifestyle changes made during the pandemic. The webinar featured Hayden AI CEO Chris Carson and Vice President of Engineering Vaibhav Ghadiok, both co-founders of the company, on some of these emerging trends.


Ghadiok said new technology such as computer learning and computer vision are making it possible to explore new transportation options.


“Machine learning, especially as it is applied to computer vision, has made significant leaps in the past decade,” he said. “We are perfectly positioned to recognize the decades-long dream of smart cities. The technology is there.”


Carson said one goal for many cities is for universal basic mobility, to make public transport easy and accessible for all.


“Smart transit ecosystems in smart cities are not just about pushing the boundaries of technology but addressing social and environmental issues like accessibility and sustainability,” Carson said. “It’s not just a technical question; it’s also a cultural and societal question. If it’s not sustainable or we can’t afford for everyone to use it then that’s not an equitable solution.”


With a multitude of new technological trends emerging, Carson said it is important for transportation authorities to stay current.


“I think the approach transit agencies must take is to embrace technology, to see it as a growth opportunity and an opportunity to innovate, have a discussion with their citizens about how they can maximize the upside of that technology in their day-to-day lives,” he said. “Smart traffic management and what makes such things possible are technologies like 5G, augmented reality, and of course the internet of things. It’s not just about the internet, but about collecting data and what you do with that data. We need to be able to sense and control resources, supply, and demand.”


One of the ways Hayden AI is working with cities to solve traffic issues is through deploying computer vision cameras on public buses and other vehicles to collect data and build virtual maps of cities.


“We are essentially deploying our transit cameras in city fleets that include school buses, transit buses, and street sweepers,” Ghadiok said. “That converts them into mobile sensors that are connecting valuable, vision-based data. While these vehicles go down these streets, they are building a 3-D map of the environment. We understand here is a sidewalk, here is a lane, here are the lane lines, here are the parking meters and fire hydrants. Essentially, what we are doing is taking some of the advances in autonomous tech and applying those with our camera, allowing our devices to be fully aware of what is happening. It’s real-time situational awareness.”


These cameras can identify visual elements like license plates, centimeter-accurate locations of buses, and street signs using both computer vision and GPS. It takes about two hours for a bus to make a map of its route. Ghadiok said the cameras also have the ability to tell the difference between pedestrians, cyclists, parked cars, moving vehicles, and road signs.


“The system can also recognize traffic lights, traffic signs, and can recognize graffiti on a stop sign and alert the authorities” Ghadiok said. “We can also do traffic pattern analysis. We can see at an intersection how many pedestrians are walking across that intersection at different times of the day. Another thing is identifying sidewalk hazards.”


As a result, the technology can provide invaluable data when it comes to making decisions on handling congestion.


“For instance, you are coming to a certain intersection between four and six p.m. and it has so many people crossing the intersection that it blocks traffic,” he said. “We can look to see what we can do to modify that street and unblock traffic. We can have a full understanding of what is blocking the roadways – whether its micro mobility scooters, bicyclists, or pedestrians – once you have insights at that level, you can start working toward and collaborating with other agencies to make transit more efficient.”


One of the ways this technology is being used is to detect vehicles that are illegally parked in bus lanes. Ghadiok said cities that have deployed this technology have seen their bus speeds improve by as much as 50%, reduced passenger wait times, increased ridership, and have saved money by making public transit becoming more efficient.


“Cities like New York have a massive problem with people who come in and park their vehicles on the streets, blocking the bus lanes,” he said. “The problem was so acute that the pre-pandemic bus speeds in Manhattan were less than 5 miles per hour. That was causing people to not get to work on time, lose jobs, not get proper medical care, and spend hours waiting for a bus. That is where that AI camera can be deployed in a bus and as it goes down the street, automatically identify the vehicles that are parked in the bus lane. It has a conceptualization of all the different lanes, which is the bus lane, and an understanding of the enforcement days and enforcement times. It automatically identifies that vehicle and creates an evidence package it sends to directly on to the appropriate agency for further action.”

While the technology has the ability to tag certain common elements along the street, Carson said the system is also designed to protect privacy for members of the public.


“What our system isn’t is a surveillance system,” Carson said. “It is not constantly recording video like a fixed camera would. Those videos are post-processed down the road and saved on a DVR. That is not what we do at all. We are running algorithms using computer vision, and we don’t capture any video until we actually observe a violation. This technology protects citizens’ privacy, and I want to reinforce is not a surveillance system like fixed cameras.”


In addition to cars parked illegally in the bus lane, Ghadiok said the technology can also be used for other types of parking enforcement. Both planned traffic routes and parking enforcement applications by the cameras can be updated to accommodate changes for major events, different enforcement rules on weekends, and for other needs as they come up.


“We are also able to identify parking spaces, so we can make an impact on parking management,” he said. “We can inform other commuters that there are vacant parking spaces in an area. Almost 30% of congestion in urban downtowns is caused by people circling to find parking. That has a massive impact. We can see there is a parking meter here and a parked car. We can make the AI understand the meter has expired and jot down the license plate of the car that is potentially in violation. Our system can go down the street and identify what cars are parked legally and which cars are illegally parked.”


With deliveries and ridesharing becoming more common, Ghadiok said new technology can help meet arising traffic needs surrounding the demand for curb space.


“The curb has become the most valuable real estate in the city,” he said. “There are Ubers, Lyfts, Amazon, FedEx, and UPS drivers all competing for that curb space. They sometimes block traffic, causing even more traffic. With our technology, we can unleash a future where we can schedule curb space. Someone can be given 15 minutes here, do their business, and leave. We can also enforce it.”


Carson said the technology can also have a positive impact on government efficiency in other areas to make data-driven decisions.


“As we expand beyond transit buses to school buses, street sweepers, garbage trucks, and police cars, we can aggregate the data and share insights across multiple government agencies, empowering them with intelligence-led operational efficiency.” Carson said. “By using digital twin technology, we can map all objects that exist in a city environment to create a duplicate, virtual space that city leaders can use make sense of community challenges and quality of life issues. We can provide data for simulation modeling and actionable insights for more intelligent operations or enforcement. Some of the things we can start to understand through traffic pattern analysis and other things is where those choke points are for bus lanes, how they can improve bus infrastructure, and others.”


Rather than waiting to see what new technology brings, Carson said now is the time to start making emerging technology part of the way cities operate.


“The technology fundamentally needed to change traffic enforcement are upon us now,” he said. “In fact, President Biden is pushing for an overhaul and update to the nation’s infrastructure, calling it a transformation effort that could create the most resilient, innovative economy in the world and a once-in-a-generation investment in America. I think it is time to embrace technology to create safer, smarter, and more sustainable transit ecosystems.”