Intel’s AI platforms helped optimize logistics for the Olympic Games. What can they do for your enterprise? – NBC New York

The following content is created in partnership with Intel. It does not reflect the work or opinions of the NBC New York editorial staff. Click here to learn more about Intel.

The Olympic and Paralympic Games Paris 2024 are big events—dozens of venues, 10,500 athletes, 20,000 journalists and 45,000 volunteers. The enormous task of planning begins long before the doors open and any people set foot inside. And with the International Olympic Committee committed to making Paris 2024 the most sustainable Olympic Games ever—with half the carbon emissions of prior summer editions of the Olympic Games—it’s clear that the need to optimize logistics demands a new kind of solution. A solution that Intel technologies are poised to deliver.

To achieve a broad set of goals—dependable, flexible planning for the Olympic venues, advance placement of everything from broadcast cameras to retail kiosks, logistical support that allows for the planning of everything from supply routes and athlete services to fan concessions—Intel’s AI platform digital twinning technology has been deployed, backed up withpeople counting enabled by Intel’s AI-platform, to deliver near real-time metrics and build on the activation for future Olympic Games—and beyond.

Intel’s AI platform digital twinning: planning with virtual replicas

Digital twins are virtual representations of objects, environments or systems, based on real-world data. Given enough data, a digital twin can be used to simulate complex interactions and virtually explore ideas that would be too expensive or resource-intensive to test in reality, helping users make decisions and manage optimizations with confidence, right from the desktop.

In preparation for Paris 2024, Intel and its partners created virtual replicas, enabled by Intel® Xeon® processors, of most of the competition venues, in some cases even before construction of the venues was finalized. This enabled planners to make critical decisions about camera placements, event setup, transit links, security, crowd management and more—with the freedom to experiment, plenty of time to adjust to changing event and venue needs, and without the need to have staff on site, which cuts travel costs.

Jean-Fauste Mukumbi, Solutions Development Manager at Intel Corporation, Olympics Program Office, explains how the virtual venues were created for Paris 2024: “We start with a blueprint in digital format provided by the Olympic organizing committee, and from those files—using powerful workstations with 4th Gen Intel® Xeon® processors and Intel® Arc™ A770 GPUs—we’re able to create 3D models and then load them into our software partner’s platform. Once those models are uploaded to that platform, which runs similarly to a 3D gaming engine, that content can be streamed to client devices all over the world.”

During the planning process for the Olympic Games, the venue twins helped make it possible for all stakeholders to view and immediately work with changes to the plan, as updates to the 3D models were shared across the platform.

“We’re really getting smarter about how you’re moving people around, how you’re thinking about concessions, how you’re thinking about signage, how you’re thinking about broadcast cameras and using those digital twins to make decisions and scenario plan right around those things,” says Sarah Vickers, head of Intel’s Olympic and Paralympic Games Program, “whereas traditionally that might have had to be done in person through 2D drawings.”

Making it count: understanding customer needs on-site

Planning is one thing, but once the Olympic Games are on it takes careful measurement to understand its effectiveness, and Paris 2024 will also showcase the way Intel® AI platforms can be used to understand customer satisfaction. While the digital twinning platform used to plan Paris 2024 won’t be updated with realtime data during the events themselves, Intel AI platforms will continue to deliver critical customer insights onsite, with a people-counting system installed at venue media centers and Olympic family lounges within all of the Olympic sites in the Paris area, Lille, and at Chateauroux.

According to Mukumbi, the system uses “stereoscopic sensors along with machine learning to count the number of people coming in and out of the different venues.” This allows for optimization of resources like food and beverage supplies, security, transportation needs and more, based on realtime data about venue occupancy. As installed, the system “monitors the speed and height of an object that’s moving and based on those two elements it can decide whether it is a person or, for instance, a pram.”

As with many venues, the Olympic Games have kept track of people counts in the past, but this was formerly a manual process, not only cumbersome and less accurate, but more significantly left no record for the future to drive further optimizations. The IOC, Mukumbi tells us, was “very interested in implementing this project because it can provide historical data that can help plan future Olympic Games.” Plus, the new system is 95% accurate, making the data far more reliable for future use.

So what will journalists at the media centers and users of the family lounges notice? According to Mukumbi, “attendees will certainly notice the level of service that will be provided to them—there will be better resource allocation on site to help support them. We expect there will be a greater level of satisfaction for the people hosted by these venues.”

Virtual spaces and live data: the future of optimization

For Paris 2024, the digital twin platform will only be used for planning; there’s no on-site  application currently in development for this year’s Olympic Games. But there is huge potential for future integrations, says Mukumbi: “Real-time data can be integrated into the digital twin platform and one of the interesting use cases for the future could be to have those functionalities merged into a single application.” In future Olympics, or other sporting events, concerts, conventions, transit centers or other busy spaces, these solutions can be deployed across multiple sites, and in the future live data integration could enable optimizations and efficiencies that go well beyond what’s possible in Paris this summer.

“The technology is ready to be deployed on that scale,” says Mukumbi. “The possibilities really come down only to the type of data you want to collect on site.”

“Think about queue times,” says Vickers. “You can use data to adjust your resources on site so that people experience shorter queues at concessions, so they’re optimizing their enjoyment, you’ve got happier customers, they’re spending more money, and you’ve got a more efficient event. That’s a win-win.”

Learn more about the ways Intel’s AI platforms are powering the Olympic Games.

More from this series: