Project Aria
Last reviewed
Jun 3, 2026
Sources
10 citations
Review status
Source-backed
Revision
v1 · 1,549 words
Improve this article
Add missing citations, update stale details, or suggest a clearer explanation.
Suggest edit
CtrlK
Last reviewed
Jun 3, 2026
Sources
10 citations
Review status
Source-backed
Revision
v1 · 1,549 words
Add missing citations, update stale details, or suggest a clearer explanation.
Project Aria is an egocentric data-collection research program run by Meta's Reality Labs Research. It was announced on September 16, 2020 [1][2]. The program centers on a pair of sensor-equipped research glasses, also called Project Aria, that Meta describes explicitly as a research tool rather than a consumer product. The glasses have no display and show nothing to the wearer; their purpose is to record first-person (egocentric vision) multimodal sensor data that can be used to develop the machine perception and AI needed for future augmented reality glasses and contextual AI assistants [2][3].
Aria is worn by trained Meta researchers and contractors, and the device and accompanying software are provided to academic and commercial research labs through the program. Project Aria is distinct from Meta's consumer Ray-Ban smart glasses and from the Orion AR prototype: it is a research instrument, not a product that ships to the public [3][9].
When Meta (then Facebook) announced Project Aria, it framed the glasses as a way to figure out how all-day wearable AR could work in practice, and to develop the safeguards, policies, and social norms that such devices would require [1][2]. The first-person data the glasses capture was intended to feed research on machine perception, scene understanding, and the construction of a persistent spatial map of the world that Meta called LiveMaps [2].
The announcement was explicit that the glasses are not a consumer device and carry no display. As the launch post put it, the glasses "won't display any information on the inside of the lens, and research participants cannot view or listen to the raw data captured by the device" [2]. The breakthroughs the program targets, including enhanced audio and visual input, contextualized AI, and a lightweight frame, are presented as prerequisites for AR glasses rather than features of Aria itself [1].
The first-generation Aria glasses weigh about 75 grams and carry a multi-sensor suite for capturing synchronized egocentric data [4]. All cameras and other sensors are calibrated and timestamped on a shared clock with nanosecond resolution, which lets researchers fuse the streams precisely [4]. The glasses do not run AR applications; they record and securely upload data for offline research.
The table below lists the Gen 1 sensor suite as documented by Meta [4].
| Sensor | Specification |
|---|---|
| RGB camera | Rolling shutter, up to 2880x2880 resolution at 30 fps, ~110 degree field of view |
| Mono scene (SLAM) cameras (x2) | Global shutter grayscale, 640x480, 150 degree horizontal field of view |
| Eye-tracking cameras (x2) | Global shutter, 640x480, up to 90 fps, inward-facing |
| Inertial measurement units (x2) | Operating at 1000 Hz and 800 Hz |
| Magnetometer | 10 Hz |
| Barometer | 50 Hz |
| Microphones | Seven-channel spatial array, 48 kHz sampling |
| Positioning | GPS receiver, Wi-Fi and Bluetooth beacon sensing |
A separate phone companion app and desktop tools handle device control, secure upload, and access to the recorded data.
Because the glasses record video, audio, and location in public and private spaces, Meta built in several safeguards described at launch [2]. Each device shows a prominent white recording light so that people nearby can tell when data is being collected, and it has a physical mute button that stops data capture when pressed [2]. Wearers are trained on where and when use is appropriate, and where it is not [2].
Data is moved using encryption and a secure ingestion pipeline. Before researchers can access recordings, the data is automatically processed to blur faces and vehicle license plates, an anonymization step intended to protect bystanders [2]. The initial deployment was limited in scale: roughly 100 Facebook employees and contractors, primarily in the San Francisco Bay Area and Seattle, made up the first group of wearers [2].
From the start, Project Aria was tied to outside research. Meta began a pilot with Carnegie Mellon University's Cognitive Assistance Laboratory to build 3D maps of indoor spaces such as museums and airports, work aimed at AR-based navigation aids [2]. Over time the program expanded into a broader model in which Meta supplies the glasses, software development kits, and machine perception services to qualifying academic and commercial labs, which then run their own egocentric studies and publish results [3][7].
A major output of Project Aria has been a set of open research datasets recorded with the glasses, distributed through projectaria.com and the open-source Project Aria Tools. These give the wider computer vision and robotics community egocentric data with high-quality ground truth that is otherwise hard to obtain. Key datasets are summarized below [5][6][7].
| Dataset | What it contains |
|---|---|
| Aria Everyday Activities (AEA) | 143 daily-activity sequences from multiple wearers across five indoor locations, with globally aligned 3D trajectories, scene point clouds, per-frame 3D eye gaze, and time-aligned speech transcription [5] |
| Aria Digital Twin (ADT) | 200 sequences in two fully digitized indoor scenes with 398 object instances (324 stationary, 74 dynamic), plus 6DoF device and object poses, 3D human poses, eye gaze, segmentations, depth maps, and photorealistic renderings [6] |
| Aria Synthetic Environments (ASE) | 100,000 procedurally generated apartment-scale interior scenes for large-scale scene-understanding research [7] |
| HOT3D | A benchmark for vision-based understanding of 3D hand-object interactions [7] |
These datasets cover everyday scenarios such as cooking, exercising, and playing games, and they pair raw sensor streams with the kind of ground truth (precise poses, geometry, gaze) that supports training and evaluating perception models [5][6].
Project Aria connects to Meta's wider egocentric research effort. The glasses are the capture device for the egocentric portion of Ego-Exo4D, a large multimodal dataset that records the same activity from both first-person (egocentric) and surrounding third-person (exocentric) cameras at once [8]. Ego-Exo4D was built by Meta's FAIR group, Project Aria, and roughly 15 university partners, and its first-person video comes with the Aria glasses' seven-channel audio, IMU data, and grayscale SLAM camera streams alongside the RGB feed [8].
Ego-Exo4D followed the earlier Ego4D effort, a consortium dataset released in 2021 by FAIR and more than a dozen universities. Together, Ego4D and Ego-Exo4D, with Aria as the hardware backbone for the newer collection, have become widely used resources in egocentric computer vision and embodied AI research [8].
On February 27, 2025, Meta announced Aria Gen 2, a second-generation research device [3]. Like the original, Gen 2 has no display and is not sold to consumers; it is aimed at researchers working on machine perception, egocentric and contextual AI, and robotics, and Meta opened it to academic and commercial labs through Project Aria, with availability to outside researchers beginning in 2026 [3][9].
Gen 2 keeps a lightweight frame, weighing between 74 and 76 grams depending on size, and adds folding arms for easier storage and transport [3]. Meta cites six to eight hours of continuous use per charge [3]. The sensor suite is upgraded: the device carries four computer-vision (SLAM) cameras, double the two in Gen 1, for a wider field of view and better stereo coverage, along with one RGB camera and two eye-tracking cameras [3][10]. The RGB/CV global-shutter sensor offers a high dynamic range of 120 dB, up from about 70 dB in Gen 1, and the stereo overlap between the CV cameras grows to 80 degrees from 35 degrees, improving depth perception [10].
Gen 2 also introduces two sensors embedded in the nosepad: a photoplethysmography (PPG) sensor that can measure the wearer's heart rate, and a contact microphone that helps separate the wearer's own voice from that of bystanders [3][10]. The device retains an IMU, barometer, magnetometer, GNSS positioning, and a spatial microphone array, and it runs SLAM, eye tracking, hand tracking, and speech recognition on-device using Meta's custom silicon [3][10]. For scale, Meta has noted that Gen 2 (about 75 grams) sits between the roughly 50-gram consumer Ray-Ban Meta glasses and the heavier Orion AR prototype, underscoring that Aria is a research instrument separate from both [9].