IT Brief India - Technology news for CIOs & IT decision-makers
India
Ropedia launches Xperience 2 dataset for Physical AI

Ropedia launches Xperience 2 dataset for Physical AI

Thu, 19th Nov 2026 (Today)
Joseph Gabriel Lagonsin
JOSEPH GABRIEL LAGONSIN News Editor

Ropedia has introduced Xperience 2, a dataset for robotics and embodied AI intended to address a data bottleneck in Physical AI development.

The product links first-person recordings of human activity with 3D reconstructions of the environments where those actions take place. That structure allows the same source material to be reused for simulation, robot training trajectories and evaluation without relying on physical robots at every step.

Robotics groups have faced a persistent challenge in gathering large volumes of useful training data for machines expected to operate in homes, shops, offices and other real-world settings. Collecting data through physical robots can be costly and slow, while many existing datasets separate human behaviour from the spaces in which it occurs.

Xperience 2 is designed to combine those elements. The dataset records human activity through the company's HOMIE head-mounted device and aligns those recordings with multimodal data and reconstructed 3D scenes, allowing developers to track both an action and its physical context.

The current dataset includes 1,812 reconstructed scenes from 596 collection locations and 4,230 task recordings. According to Ropedia, the material covers activities including object manipulation, cleaning and household tasks across settings such as homes, restaurants, retail stores, offices and outdoor spaces.

Four routes

Ropedia said the dataset can be used in four main ways. The first is real-to-simulation reconstruction, in which individual objects scanned in the real world are turned into interactive simulated assets with collision properties, joints and physical parameters.

In one example, a single home produced 107 objects for simulation, 73 of which could be opened, pressed or deformed. A second route translates recorded human movements into robot joint trajectories, allowing demonstrations to be adapted for different robotic systems.

The third route uses virtual teleoperation to generate robot interaction demonstrations in simulated settings, reducing the need for manual resets and repeated physical testing. The fourth produces long-horizon robot trajectories, enabling models to practise multi-step tasks involving several objects or rooms.

One demonstration described by Ropedia involved a simulated wheeled dual-arm robot carrying out eight consecutive subtasks in a reconstructed environment. Those tasks included retrieving a cup, operating a microwave and placing the cup in another room.

All 30 executions in that test were completed across different starting configurations without intermediate resets, according to the company. Each trajectory lasted about 2.2 minutes, while generation took 10 to 16 minutes on a single simulation worker.

Development loop

Beyond data generation, Ropedia is positioning Xperience 2 as part of a broader development and evaluation process for robotics teams. The company said researchers can use consistent tasks and starting conditions to compare models, measure completion rates, identify failure patterns and determine where more data is needed.

The process also allows changes in object positions, scene layouts and robot configurations to test how systems behave under different conditions. The aim is to create a feedback loop between simulation results, robot performance and future data collection.

Ropedia traces the launch to its wider effort to build data infrastructure for what it describes as Physical AI. Its broader Xperience collection now spans 10 million real-world interaction clips, or about 100,000 hours of data, and has recorded more than 2.7 million downloads on Hugging Face, the company said.

The business operates from four offices worldwide and works with more than 20 robotics and foundation model companies. Xperience 2 sits alongside its HOMIE capture hardware, Foundry processing platform and data services business as part of that offering.

"To scale Physical AI, we need infrastructure that makes every real-world experience more valuable by connecting it to the environments, actions, and outcomes that robots need to learn from. With Xperience 2, we are building towards a data infrastructure where human experience, 3D environments, simulation, and evaluation work together, allowing a single demonstration to support multiple training and development workflows," said Zhaoxi Chen, Co-Founder and Chief Executive Officer, Ropedia.