Three Anzellan robots can see, hear, and react to guests through AWS, but every movement and line of dialogue was scripted in advance by Disney's Imagineering team.
- Walt Disney Imagineering R&D has unveiled three experimental Audio-Animatronic figures: Vilo Vele, Kado Bre'a, and Jibu Bre'a
- Hidden cameras and microphones capture guests' video and speech, which AI models running on the AWS cloud then process
- Imagineers scripted every movement and line in advance, and the AI only selects and triggers them at the right moment
- They will debut publicly at the AWS re:Invent conference in Las Vegas, and could later appear in Star Wars: Galaxy's Edge
Table of Contents
Three Anzellan robots make their debut
However sophisticated and lifelike Disney's park robots have become, most still run on pre-programmed motion, or on a hidden operator who makes their behavior look more natural. That may soon change: Walt Disney Imagineering R&D has just announced three experimental Audio-Animatronic figures that use the AWS cloud and Amazon's AI capabilities to respond differently to each guest in real time.

The robots are named Vilo Vele, Kado Bre'a, and Jibu Bre'a, and are based on the Anzellan alien characters from Star Wars that appeared in The Mandalorian and Grogu, released earlier this year. Each has its own personality. Their first public appearance will take place in Las Vegas, where attendees of the AWS re:Invent conference can chat with them. In the longer term, the characters and the technology behind them could move on to Star Wars: Galaxy's Edge and other attractions in Disney parks.
How the system recognizes guests
Unlike the recently unveiled Olaf robot, which needs someone quietly operating it with a Steam Deck handheld, the Anzellans have no one pulling the strings. Each robot is fitted with hidden cameras and microphones that capture guests' movements and speech, and multiple AI models running on AWS cloud servers then process that data.
The models split the work by how fast a response is needed. A "fast-decision model" handles instant reactions, such as quickly meeting a guest's gaze so they feel the robot is paying attention to them, while larger models handle longer interactions.
The dialogue is still written by humans
Even so, the Anzellans' movements and replies are not improvised on the spot by AI. Every gesture and line that the models select and trigger was written in advance by Disney's Imagineers, to keep the animation true to each character, including each one's individual quirks.
In other words, the AI helps steer the robots' actions, but what they say and do never goes beyond the script.
Disney's approach is fairly pragmatic: let AI handle seeing, hearing, and choosing when to react, while the content stays in the hands of human writers, so the characters are unlikely to drift from their established personalities in front of a big crowd. That is why "AI interaction" is best understood as AI acting as the one who picks the scene, not an actor improvising, and it is also why this direction looks safer for a theme park. For Vietnamese readers, this is first and foremost a signal of how major parks will build robots in the coming years, since there is no information yet on when this technology will reach real attractions. If you are planning a trip to Disney, there is no need to change your plans because of this; think of it as something you may encounter at Galaxy's Edge down the line.
Source: The Verge
Reviewed by: Hero Association








