Meet ADAM, the Nvidia-Powered NHL Arena Bartender Robot

The hospitality sector is increasingly embracing robotics to ease persistent labour shortages, with rollouts now extending beyond kitchen automation into guest-facing roles.

A case in point is the T-Mobile Arena in Las Vegas, where hockey fans at Golden Knights games were served drinks by ADAM, a robotic bartender that demonstrates automation can manage more than just repetitive tasks.

What is ADAM’s purpose?

ADAM – short for Automated Dual Arm Mixologist – is a real-world deployment of edge AI computing in hospitality settings.

Created by Las Vegas-based Richtech Robotics and built on Nvidia’s Isaac libraries, the system tackles staffing pressures while enabling what the company describes as distinctive customer interactions at the NHL venue.

“The hospitality industry faces significant labor challenges and ADAM is our answer to meeting those needs while elevating the customer experience,” says Matt Casella, former President of Richtech Robotics as of 2nd December this year. 

“With Nvidia’s Isaac platform, we’ve developed a solution that’s scalable, consistent and frankly, creates memorable moments for fans.”

Before ADAM served its first drink at the arena, it spent substantial time training in a virtual bar.

Richtech used Nvidia Isaac Sim, a robotics simulation framework built on Nvidia Omniverse, to create a digital twin of ADAM’s workstation.

The simulated environment included cups, utensils and varied lighting conditions that could impact vision systems, enabling the robot to master real-world challenges before encountering them.

How was ADAM created?

This simulation strategy produced synthetic data – artificially generated training inputs that teach AI systems to recognise objects under varying conditions.

It proved especially effective in training ADAM to identify items even when confronted with glare or reflections that could confuse camera-based vision systems.

Isaac Lab, Nvidia’s open-source robot learning framework, then honed ADAM’s operational skills, including pouring and shaking beverages.

The result is a robot that goes beyond fixed programming, adapting to its environment with precision.

ADAM runs on Nvidia Jetson AGX Orin, an edge AI computing platform delivering up to 275 trillion operations per second of performance.

Edge AI processes data locally on the device rather than sending it to the cloud, reducing latency and enabling faster response times.

The Nvidia technology behind ADAM’s capabilities

Leveraging Isaac ROS 2 libraries, ADAM ingests camera feeds, detects objects, and calibrates its workspace in real time.

Its perception stack, built with TAO Toolkit and optimised with TensorRT, recognises cups, measures liquid levels and adjusts movements with latency below 40 milliseconds.

In practice, ADAM can spot a misplaced cup, detect when foam reaches the rim, and correct a pour without missing a beat.

Beyond hospitality, Richtech has developed Dex, a mobile humanoid robot built for factory and warehouse applications.

Unveiled at GTC DC, Nvidia’s technology conference, Dex pairs an autonomous wheeled platform with dual‑arm manipulation for tasks including machine operation, parts sorting, and material handling.

The system runs on Nvidia Jetson Thor – a robotics processor designed for real-time sensor processing in industrial environments – and was trained using a blend of real-world and synthetic data generated with Isaac Sim.

“The response at T-Mobile Arena has been phenomenal – people love interacting with ADAM,” Matt says.