KMWorld 2024 Is Nov. 18-21 in Washington, DC. Register now for Super Early Bird Savings!

BI and ANALYTICS sustain smart city initiatives

Article Featured Image

There are many opportunities for improving pedestrian safety through advanced technology. “Crosswalks have not changed in 50 years,” Finochio pointed out. “People push a button and wait to cross the street.” With better technology, it will be possible to collect more information and use it to adjust crossings. “We can monitor whether the person is still in the crosswalk before the traffic light changes back to green, and we can determine whether the individual is in a wheelchair. We can also help drivers focus on the real-time information instead of static signage. Many intersections have a lot of visual pollution, with numerous signs and moving objects. Through these projects, it will become feasible to send an alert to drivers via smart devices, the vehicle’s dashboard, or by triggering an immediate LED warning message using a sign while a vehicle is approaching an intersection.”

Some of the cameras used in the Denver system are DNNCams from Boulder AI. They are designed for harsh environments and have built-in neural intelligence that extracts data using AI techniques such as deep learning to interpret the data. The data is then sent to the cloud, where it is analyzed further. It includes counts of cars, number of near-misses, and other measures that can help determine how dangerous an intersection is. Unlike traffic cameras that have the single purpose of providing visual information, the DNNCam collects multiple measures about activity in the intersection.

“The cameras take the place of purpose-built traffic counters or human observers, who may only be able to be in the field a few times a year. In that situation, it takes many studies to quantify the safety of an intersection,” said Darren Odom, founder and CTO of Boulder AI. “In addition, our methodology is very privacy-oriented. By sending just the metadata to the cloud, we do not collect personally identifiable information. Cities are becoming increasingly sensitive to the issue of privacy, and we take extraordinary measures around the way we transmit data.”

A platform approach to citizen services

Although many cities begin with a point solution and then expand to additional applications and functionality, the lack of interoperability can create barriers that prevent the siloed data from being shared. Purpose-built smart city platforms facilitate an integrated approach to collecting and analyzing data across the city’s applications. Lumada is an IoT solution from Hitachi Vantara that supports the creation of smart space applications for city infrastructure, safety, and transportation.

In India, the state of Andhra Pradesh, which has a population of 50 million, has created a real-time governance center and made all its services data-driven using smart spaces technologies. For example, analytics allows the government to improve distribution of food rations, reduce pension fraud, and notify citizens when a storm is coming. These applications are all integrated on a common platform that allows for analytics across the various datasets. “There is a huge ROI on this type of initiative,” said Justin Bean, global director of Smart Spaces marketing. “Savings in resources such as food, in dollars, and in time are all occurring.”

One way for cities to move forward with smart city programs if infrastructure is limited is to integrate with existing private sector technology. “Many companies have security cameras, for example,” noted Bean, “and they are often willing to share with local public safety officials when crimes occur. These can be seamlessly integrated with municipal systems. Clearly, in criminal cases, the availability of video is a big asset.”

In order to protect privacy, only authorized law enforcement personnel are able to access the video, and their usage of the data is auditable to provide transparency and ensure compliance with privacy regulations. Privacy is also protected by the use of intelligence on the edge and by masking through pixelation.

KMWorld Covers
for qualified subscribers
Subscribe Now Current Issue Past Issues