Gesichtserkennung bei Attraktionen

SIX YEARS AGO, CALIFORNIA-BASED AUDIOVISUAL solutions provider Mad Systems began exploring ways to provide greater personalized content for museum visitors. How could they deliver customized information at an exhibit, based on a guest’s age, language, and interests? Initially, Mad Systems considered barcodes, RFID wristbands, and QR codes, as well as a custom-made infrared badge. Several shortcomings, though, quickly emerged. For one, devices had to be recharged. Frontline employees would also need to sterilize and properly account for the equipment at the end of a guest’s visit.

“We got to a point where we realized that we needed to find something that we didn’t have to clean or charge and where we didn’t care if people took it home with them,” explains Mad Systems founder Maris Ensing. “Then it became obvious: The best tool is a person’s face. It’s unique and we don’t have to look after it.”

Like others working in this space, as Ensing went down this path he discovered that artificial intelligence-driven advances in facial recognition promise to reshape how attractions engage with guests—and potentially usher in a new era of greater convenience, personalization, and security.

At the Front Gate

Convenience and security were among the driving factors at Changi Airport’s Terminal 4 in Singapore, where cameras with facial and iris recognition technology have replaced immigration officials and cut the time required to clear international travelers by more than half. The technology is also used for check-in, baggage drops, and boarding. Millions of passengers have used the system since it was introduced late last year.

Universal Destinations & Experiences was also an early adopter of facial recognition technology. Photo validation is now an option at the entrances to Universal Studios Florida, Universal Islands of Adventure, and Universal Studios Beijing to provide a faster, more seamless experience for visitors.

“Upon entering the theme parks, guests are asked to take a photo instead of scanning their fingerprint. The image is then securely stored and, on subsequent visits, is used to validate admission,” explains Universal Destinations & Experiences’ Director of Corporate Communications Brian Bruce.

At Universal Studios Beijing, photo validation is also used by guests to rent lockers. The resort’s website notes that facial recognition shortens waiting times and contributes to a lower carbon footprint by eliminating paper tickets.

“Photo validation is completely optional, and guests can choose to opt out of the service at park entry or by visiting guest services before, during, or after their first visit,” Bruce adds. Photos at Universal Studios Florida and Universal Islands of Adventure are deleted within six months of a guest’s visit to the park. Universal Studios Beijing retains digital images and templates for up to one year.

Bruce shares with Funworld that Universal Orlando is preparing to expand its use of facial recognition technology. The operator has already installed cameras at the entrance to Universal Express lanes for several attractions, including the Trolls Trollercoaster in DreamWorks Land at Universal Studios Florida. The biometric system will identify guests and prevent unauthorized access.

Records show Universal has also filed a patent for a machine learning system that uses facial recognition along with temperature, sound, motion, and other sensors to monitor and improve the overall guest experience. In addition to monitoring line jumping and unsafe movement on rides, the Amusement Park Interactive Guidance System can reportedly track liquid consumption and determine if a guest should be flagged from purchasing additional alcoholic beverages. Universal submitted an updated patent application for this technology in 2024, after originally applying for patent protection in 2022.

Privacy Issues

While facial recognition has a broad range of use cases in the attractions industry, operators need to anticipate some guests will remain wary that their data could be shared or stolen, potentially leading to unauthorized surveillance, identity theft, or third-party tracking. Attractions that adopt this technology need to address these privacy and data protection concerns directly.

“We recognize that the public has significant concerns regarding data privacy, ethics, and the social implications of facial recognition technology, resulting in a cautious approach to its integration,” says Ensing.

“Given these considerations, our patented CheshireCat solution operates within a controlled environment,” he explains. “The server does not connect to the internet and it can be put in a locked room. Thus, ensuring it’s secure. This mitigates potential data breaches and external access. The data is not shared or accessible outside of the intended context.”

User information is also scrubbed from the CheshireCat databases daily to ensure no personal data is stored long-term.

Once Mad Systems decided to proceed with a server on-site, they discovered an important secondary benefit: the system operates significantly faster. Unlike remote servers which can experience latency times, CheshireCat recognizes individuals within a fraction of a second.

“Imagine that you’re standing in front of an exhibit and it takes five or more seconds for the computer to know that you are there,” says Ensing. “That doesn’t work. Visitors are losing interest and walking off to another exhibit at that point.”

Customization

While Universal Destinations & Experiences’ current facial recognition efforts are focused on entry points, Mad Systems plans to use the technology to deliver personalized content to museum visitors.

Imagine being able to offer translations in hundreds of languages—via audio or machine-generated subtitles—and tailor an exhibit’s description based on a person’s age and interests. Young children can listen to one explanation with age-appropriate vocabulary, while teenagers are delivered another version. Adults might tailor their content based on whether they are a “stroller” who likes more detail or a “grazer” who moves more quickly. Similarly, if an exhibit is about Leonardo da Vinci, guests could choose whether they are most interested in learning about his art or science.

At the base of Mad Systems’ work is a non-proprietary audio-video system named QuickSilver, which can run hundreds of parallel audio tracks—hence the ability to offer multiple languages and styles. Built on top of that is an artificial intelligence system named Alice that creates personalized content; a distribution system named Lory that delivers the content to personal devices; and the facial recognition system CheshireCat, which matches the content to each visitor based on their preferences and demographics.

Audio content can also be delivered directly to hearing aids, while translations in sign language can be sent to a visitor’s smartphone.

Mad Systems is also at work on customized avatar docents. “When you walk in, there are four virtual people: a hydrogeology engineer, a research assistant, a field worker, and an administrator. Each avatar can guide you through the exhibition, providing a different perspective,” says Ensing, whose portfolio includes projects at Mattel Adventure Park, Crayola IDEAworks, and the Truman Presidential Library and Museum.

Is Facial Recognition for Animals Next?

Scientists have developed facial recognition tools for wildlife and livestock to track individual animals, but the technology is not yet widely adopted by safari parks.

Instead, companies like Cincinnati-based Pigeon Tech have deployed image recognition and gamification to enable visitors to identify animals in real time is in use at the Fossil Rim Wildlife Center in Texas and in France’s Parc Animalier d’Auvergne. Visitors to the Museum of Flight in Seattle can use the technology to identify planes. Like Mad Systems’ solutions, the program does not require an internet connection.

“Guests can point their camera at a moving animal, six to eight meters away from their vehicle, and it will recognize the animal with a percentage clarity,” says Pigeon Tech Founder and CEO Vijay Sabbineni. Visitors can bookmark the animals they’ve seen and take quizzes to learn more.

“We have used it for the purpose of identifying the species,” adds Pigeon Tech Chief Technology Officer Sreenadh Kondru. “We haven’t tried it out at the animal level.” But while Pigeon Tech’s solution does not currently differentiate between the faces of two animals within the same species, Kondru says “it may be possible.”