Austin draws people to work, music venues, comedy clubs, barbecues, and more. However, along with this boom came the metropolitan depression of traffic jams.
Rekor, a traffic management and public safety analytics company, is on the front lines of dealing with increased traffic due to the influx of new residents moving to Austin. To alleviate roadway concerns, Rekor is working with the Texas Department of Transportation on a $7 billion project to address this issue.
“Texas is responding to road growth and demand by investing heavily in infrastructure, with a focus on digital infrastructure,” said Sherbin Esfahani, vice president of global marketing and communications at Rekor. . “They realized that this is so complex that traditional systems can’t really manage and understand it in real time.”
Columbia, Maryland-based Rekor uses NVIDIA Metropolis for real-time video understanding and NVIDIA Jetson Xavier NX for edge AI in Texas, Florida, Philadelphia, Georgia, Nevada, Oklahoma and many other U.S. destinations, as well as Israel. I have been using modules. and elsewhere internationally.
Metropolis is an application framework for smart infrastructure development using vision AI. Provides developer tools such as NVIDIA DeepStream SDK, NVIDIA TAO Toolkit, pre-trained models from the NVIDIA NGC catalog, and NVIDIA TensorRT. NVIDIA Jetson is a compact, powerful, energy-efficient accelerated computing platform used for embedded and robotics applications.
Rekor’s work in Texas and Philadelphia to help improve road management with AI is the latest development in an ongoing road safety and traffic management story.
Reducing rubbernecking, pileup, fatalities, and jams
Rekor offers two main products: Rekor Command and Rekor Discover. Command is an AI-powered platform for traffic management centers that quickly identifies traffic events and zones of concern. This provides real-time situational awareness and alerts to transportation departments, enabling them to keep city roads safer and less congested.
Discover leverages Rekor’s edge systems to fully automate the capture of comprehensive traffic and vehicle data, delivering robust traffic analytics that turn road data into measurable, trusted traffic knowledge. Rekor Discover gives transportation departments a complete picture of how vehicles move on the road and the impact they have, so they can better organize and execute future city-building efforts. It will be.
The company has deployed commands throughout Austin to help detect problems, analyze incidents, and respond to road activity with real-time views.
“Every time an accident occurs and stays on the road, it creates a four-minute traffic jam, puts stress on the road and significantly increases the likelihood of secondary accidents, such as rubbernecking accidents.” Paul Mathews・Mr. Zamsky stated. , Vice President of Strategic Growth and Partnerships at Rekor. “Austin implemented Rekor Command and saw a 159% increase in incident detection and was able to respond to those incidents eight and a half minutes faster.”
Rekor Command ingests many data feeds, including traffic camera footage, weather, connected vehicle information, construction updates, and leverages third-party data as well as other data infrastructure. AI is then used to establish connections and surface anomalies such as roadside accidents. That information is presented to the traffic management center in a workflow for consideration, confirmation, and action.
“They’re seeing it, they’re responding to it, and they’re doing it faster than ever before,” Esfahani said. “This not only saves lives on the road, it also improves people’s quality of life, allowing them to get home faster and out of traffic, reducing the strain on the City of Austin’s system.”
In addition to leveraging NVIDIA’s full-stack accelerated computing for roadway intelligence, Rekor is fully committed to NVIDIA AI and NVIDIA AI Blueprints. These are reference workflows for generative AI use cases built with NVIDIA NIM microservices as part of NVIDIA AI Enterprise software. platform. NVIDIA NIM is a set of easy-to-use inference microservices to accelerate deployment of foundational models to any cloud or data center while keeping your data safe.
“Rekor has multiple large language models and vision language models running on NVIDIA Triton Inference Server in production,” said Shai Maron, Rekor’s senior vice president of global software and data engineering. Masu.
“Internally we use this to annotate data and help us optimize various aspects of our daily lives,” he said. “An external LLM helps calibrate and configure the camera in a more efficient way.”
Rekor uses NVIDIA AI Blueprints for video search and summarization to build AI agents for city services, especially in areas such as traffic management, public safety, and city infrastructure optimization. NVIDIA recently announced a new AI Blueprint for Video Search and Summarization that enables a variety of interactive visual AI agents to extract complex activity from large volumes of live or archived video.
Philadelphia monitors roads, EV charger needs and pollution
The Philadelphia Navy Yard is a tourist hub operated by the Philadelphia Industrial Development Corporation (PIDC), but there are some challenges with road management and data collection about new development in the popular area. The Navy Yard’s 1,200-acre site is home to more than 150 companies and 15,000 employees, and a $6 billion redevelopment plan there will create more than 12,000 new jobs and thousands more. It is expected that more than 1,000 residents will be drawn to the area.
PIDC called for more visibility into the impact of road closures and construction projects on mobility and how to improve mobility during key projects and events. PIDC also seeks to enhance the Navy Yard’s ability to understand the volume and traffic flow of motor carriers and other large vehicles and quantify the impact of speed reduction devices deployed on hazardous sections of roadways. Ta.
Discover insights from PIDC about additional infrastructure projects you need to deploy to manage traffic changes.
Understanding the number of electric vehicles and where they enter and exit the Navy Yard provides PIDC with clear insight into potential locations for future electric vehicle (EV) charging stations. By drawing insights from Rekor’s edge systems, built with NVIDIA Jetson Xavier NX modules for powerful edge processing and AI, Rekor Discover helps Navy Yard understand how many EVs are coming and where they are coming and going. It will allow PIDC to better plan potential sites for EVs. There are also plans to expand charging stations in the future.
Rekor Discover allows PIDC planners to create EV traffic hotspot maps by referencing data provided by an AI platform. The solution relies on NVIDIA’s DeepStream data pipeline and real-time traffic analysis using Jetson. Additionally, use NVIDIA Triton Inference Server to enhance LLM functionality.
In addition to mitigating property damage, PIDC also wanted to address public safety issues related to speeding and collisions. We use speed insights to introduce traffic calming measures when the average speed on a particular stretch of road exceeds the ideal speed.
NVIDIA Jetson Xavier NX monitors contamination in real time
Traditionally, urban planners have been able to look at satellite images to understand where pollution is, but Rekor’s vehicle recognition model running on NVIDIA Jetson Xavier NX modules can track pollution sources and help mitigate them. We have taken another step towards this goal.
“Air quality is key,” said Shobhit Jain, senior vice president of product management at Rekor. “We’ve built a model that allows us to do that very well. We can tell you how much pollution each vehicle is emitting.”
Looking ahead, Rekor is exploring how NVIDIA Omniverse can be used for digital twin development to simulate traffic mitigation through various strategies. Omniverse is a platform for developing OpenUSD applications for industrial digitization and generative physics AI.
Developing digital twins for local governments using Omniverse will have a significant impact on reducing traffic, pollution, and traffic fatalities. We believe that Rekor is very beneficial to our customers in all areas.
“Our data model is very granular and we’re definitely looking at omniverse,” Jain said. “We want to look at how we can support these digital use cases.”
Learn about NVIDIA AI Blueprints for building AI agents for video search and summarization.
