Transforming Urban Cores into Smart Cities with Mobility Data and AI

Graphic of parking spaces Transforming Urban Cores into Smart C

Transforming Urban Cores into Smart Cities

Becoming a Smart City

For the first time in history there are more people living in cities than in rural areas; this trend is accelerating at breathtaking speed in China and India. The UN predicts that by 2050 two-thirds of the world’s population will live in an urban area. That adds up to nearly seven billion people – just one billion less than are inhabiting the entire planet today.

For today’s urban centers to accommodate this unprecedented growth, major changes will have to occur in the management and delivery of citizen services. These changes are driven in part by changes in mobility and the increasing use of the Internet of Things (IoT) to monitor the Built Environment. The flow of digital information may have the potential to overwhelm existing information networks.

In order to be considered responsive by the public, time frames will have to shift from being measured in months to days and eventually to seconds. Cities will have to take advantage of digital technologies such as the cloud, AI, 5G, autonomous vehicles, digital twins, robots, drones, and even blockchain, just to name a few. Doing so will transform them into resilient Smart Cities that can operate efficiently in real time and support the health, safety, and livelihood of its residents in a cost effective manner.

One of the key information components upon which virtually all of the Smart City mobile apps will be based is dynamic and timely street level data. Static, two-dimensional map layers locked in a GIS only accessible to a few highly skilled technicians will be replaced by dynamic, 3D mobile applications that provide situational awareness information. This information can then support the intelligent management of robo-taxis, infrastructure, drone package delivery, and much more with street level analytics in near real time.

A company on the forefront of the Smart City transformation is Allvision. Based in Pittsburgh, PA, this geospatial data analytics company is leveraging the robotics and AI community in the “East Coast Silicon Valley” to develop an integrated platform that fuses both photogrammetric and SLAM-based 3D lidar data. By aggregating multiple sources of geospatial data, Allvision’s platform, leveraging machine learning and advanced classification, can provide actionable information from huge amounts of reality data. This data can lead to a more informed decision making process by creating and updating a high resolution 3D map of the city including key assets.

Let’s take a look at how they are becoming a leader in supporting the development of Smart Cities.

It’s All About the Data
Since the early days of GIS, managing data and converting it into actionable, reliable information as quickly as possible has always been the primary challenge for municipal governments. Things happen fast in the Built Environment as compared to ecological processes, and the information systems found in most cities today were not designed for the kind of urban mobility challenges they are suddenly facing. Allvision brings value to burgeoning Smart Cities by introducing robotics to a world that today is used to people with clipboards or mobile devices walking along streets and sidewalks manually recording things like telephone poles, street signs, and parking zones. By digitizing the cities that we live in, optimization of curb spaces and beyond is within reach.

Asset Map Transforming Urban Cores into Smart C

Asset Map

The development of 3D mapping technologies now includes the fusion of mobile lidar sensors and higher resolution cameras that are mounted on a variety of vehicles and platforms. With these technologies, it is possible to collect a tremendous amount of raw, 3D reality data in a very short period of time. Extracting the needed 3D asset information in a timely manner requires the use of a machine learning AI platform combined with cloud computing that intelligently converts geospatial data into actionable information. This information ultimately supports timely, better informed decision making.

The Allvision machine learning platform is capable of being trained to support this kind of high-speed data processing with tremendous flexibility. With a vision of utilizing base data within different markets, the software identifies areas of change from the initially created 3D “digital twin” model.

This 3D model can be mined to track any number of street level assets that the city would like to monitor. Assets can include signage, bus shelters, street furniture, lighting, hydrants, and up-to-date parking regulations. The platform can also be applied to transit, rail, and utility corridors.

Curb Space Management is Here

Forward-looking cities are beginning to realize that their curb space is too valuable to utilize on parked vehicles alone. In the urban core, the curb is where the action takes place. If cities are able to manage curb space efficiently, they can control the flow of people, vehicles, and services in the urban core while optimizing the revenue that is generated from the usage of the curb. Curb space management is not just about parking; neither is Allvision.

The rapid rise of on-demand mobility services such as rideshare, robo-taxis, scooters, bikes, and more are driving the need for real time curb usage information. This allows dynamic adjustments in the management of valuable city resources and improves the overall user experience.

Allvision foresees many opportunities for autonomous vehicles to collect street level data while they are completing their primary task. Public works vehicles that are collecting trash or street cleaning can also acquire data and transmit it back to the data center, essentially for free. Buses, taxis, and autonomous vehicles that support the on-demand economy can also be tasked with data collection.

image of truck Transforming Urban Cores into Smart C

Managing the Curb

In order for curb management apps to be effective, mapping information needs to be available as fast as possible. This is where Allvision machine learning technology comes into play. It is designed and trained to identify objects of interest and changes from the desired state in as near real time as possible, thereby facilitating dynamic changes to the system.

The management of ADA compliance can also be dramatically improved with a dedicated curb space management program. The availability of up to date street level data could change the lives of people with disabilities as well as make sure that cities are compliant to ADA regulations.

Municipal governments are not the only customer for this information. The private sector has a number of workflows that will benefit, including construction, ride sharing, life safety, insurance, advertising, security, and package delivery logistics. Near real time street analytics would also dramatically transform a city’s public works and utilities. It also has the potential to positively affect the coming rollout of 5G networks by mapping potential antenna locations.

Validating the Business Model
In order to demonstrate their complete solution to future customers and potential business partners, Allvision is planning to embark on a series of mobile data collections around the country. Allvision will target medium-sized urban areas with their internally developed, unique, mobile data collection system that leverages several groundbreaking components from Allvision’s hardware partners.

The GIS-grade (+/- one foot) 3D mobile data collection system is highly portable, self-contained, and can be up and running within minutes of hitting the ground. Since the system is using SLAM algorithms to derive the location of the scanner, it does not require full time GNSS network support and can thus be leveraged in GPS-deprived urban canyons without the need for an expensive IMU. The software stack that Allvision has developed to process the raw data from their mobile mapping system was intentionally designed to process the type of fused lidar and image data that will be coming from autonomous vehicles of the future. These vehicles will need similar real-world data for training the AI-based driving simulators that will be required to establish safety standards for the industry.

It is important to note that Allvision does not intend to be in the mobile data collection business once the complete model has been proven. This is a short term strategy to raise awareness and start conversations with an initial set of target customers and potential franchise partners. The goal is to capture 50 to 100 miles of urban street data in 10 to 15 cities this fall.

Once these initial pilot projects are completed, Allvision plans to focus its business on a subscription model that will utilize their proprietary machine learning and cloud platforms to provide street level, geospatial data analytics to a variety of customers. This subscription is offered at a simple price-per-mile cost structure.

Mobile Data Collection Transforming Urban Cores into Smart CThis mobile data collection platform can also be used in rail and transit applications. The Allvision Virtual Rail Inspector, or VRI model, can enable automated asset tracking along rail lines by placing sensors on high rails and locomotives. Allvision has been working with a Class 1 railroad to demonstrate the effectiveness of the data collection system along with the automated recognition/tracking of physical assets and creation of a digital twin.

Pilot Project
Allvision’s project goal with the City of Pittsburgh was to observe and analyze peak parking behavior between 16th and 26th street along Smallman and Penn Avenue and select side streets in the city’s Strip District to better understand current usage of the curb and parking infrastructure, particularly understanding drop-off/pickup and delivery patterns. The end report detailed parking usage classified by civilian and commercial use along with vehicle type (motorcycle, car, truck, construction, etc.). Areas of under/over utilization as well as illegal usage were also identified.

Over two consecutive weeks on Tuesdays, Thursdays, and Sundays from 11 AM to 2 PM, the total capture counts were as follows: 59 faces, 314 spaces, and 11,719 vehicles over 121 miles and 639,516 feet. The study was able to reveal key learnings on construction’s impact on revenue, which streets are in high demand for parking spaces and when, and the major types of parking violations (mostly loading zone infractions carried out by civilian vehicles). These digital insights will be invaluable in shaping the future of the Strip District and the parking landscape of Pittsburgh in general. With Allvision’s help, Pittsburgh is on its way to a Smart City designation.

Smart Cities of the Future
Virtually every city, large and small, has begun laying the groundwork for the types of services and kinds of environments that will be required to satisfy the needs of rapidly growing urban populations around the world.

Singapore, with its unique geographic and political identity, is one of the most advanced city states in terms of implementing smart city initiatives. It is estimated that they plan to spend US$1 billion in 2019.

Singapore is at the forefront of the digital economy, digital government, and digital society. One could argue that they are providing the rest of the world a preview of what is to come. Allvision is continuing to do their part to support this important global smart city transformation.

Catch Allvision at one of their upcoming events, including INTERGEO in Germany, the IPMI Leadership Summit in Pittsburgh, PA, or the NPA Conference and Expo in Orlando, FL.

Email Allvision at to set up a meeting.

Note – If you liked this post click here to stay informed of all of the 3D laser scanning, geomatics, UAS, autonomous vehicle, Lidar News and more. If you have an informative 3D video that you would like us to promote, please forward to and if you would like to join the Younger Geospatial Professional movement click here.

One comment

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.