01
Build a focused prototype
Select an open-source arm design, print parts, assemble the hardware, and validate basic desktop tasks.
Lab & product
Understand the activity. Help with what comes next.
Matha Labs starts with datasets for human life activities from the first-person view. We’re working toward ten focused collections, then a broader library with evaluated models for developers and HumanHUD. HumanCapture is our planned contribution app; robotics concepts are preserved here as much later research.
In development / A Matha Labs product
HumanHud.ai
Your life, in a calendar and transcript.
We’re building HumanHud.ai to transcribe a person’s daily life through compatible glasses into a calendar and transcript that AI agents can use. Our integration direction includes agents such as Meta Muse, Hermes, OpenClaw, and OpenAI Dot; these integrations are planned.
Where we’re taking it
Our first ten datasets focus on everyday activities from the first-person view, such as cooking, exercise, reading documents, and computer work. HumanHud.ai is intended to turn that activity context into a useful personal record. HumanCapture is our planned app for paid, opt-in glasses-POV capture and annotation; evaluated models and developer APIs connect the shared library to future applications.
Capture the context
Use a compatible first-person view, sound, and activity context to understand the moments of daily life.
Build a personal record
Organize what happened into a calendar and transcript, so useful moments can be found again.
Give agents useful context
Build toward permissioned access to a person’s calendar and transcript so connected AI agents can help with relevant context.
Platform capabilities
Capture capabilities vary by device. The current Meta Ray-Ban Display web interface is display and D-pad only; it does not have camera or microphone access. Activity capture requires a compatible platform or companion camera, and those integrations are in development.
Later research / Robot interaction SDK
From seeing the task to guiding the robot.
A robot interaction SDK is a possible later extension of HumanHUD, after our first-person dataset work. The concept would make a compatible robot’s intended task visible, let the person guide permitted actions, and connect that interaction to useful training examples.
- 01
Select
Choose a visible, compatible robot the person has permission to operate.
- 02
Understand
See the intended task and the actions available for that integration.
- 03
Guide
Use hand gestures to issue permitted commands through the glasses.
- 04
Learn
With participation, pair the first-person view and instruction with the robot’s outcome.
Robot SDK and integrations are later research with no scheduled release.
01 / Datasets & models
Human activity. A shared intelligence library.
We’re building a library of AI models and datasets that categorize what people do. The same intelligence will power HumanHUD and be available to developers through APIs.
One library. Two ways to build.
HumanHUD brings activity intelligence to AR glasses. Our planned APIs will give developers access to the same model library for their own applications.
Later research / Robotics
From understanding to physical assistance.
Our current focus is first-person datasets for human life activities. Robotics is a longer-term direction. This interactive illustration preserves our exploration of how useful activity understanding could eventually support machines that assist people.
Much later / Preserved research concept
A future experiment. One arm at the desk.
This desk-mounted robot arm concept is preserved for much later research, after our first-person dataset work. An open-source, 3D-printable design and a fixed workspace could offer a focused setting for studying human demonstrations, permitted commands, and observed outcomes.
Later research · Unscheduled · No completed hardware prototype claimed
02
Connect HumanHUD
Prototype robot selection, visible task intent, and permitted hand-gesture commands through a compatible integration.
03
Collect useful examples
Pair demonstrations and robot outcomes, then develop simulated variations for custom enterprise datasets.
Possible later research sequence
- 01Dataset foundations first
- 02Later: desktop arm + SDK studies
- 03Later: mobile arms + robot shed
- 04Later: wearable backpack research
Preserved kit study: SO-ARM101 Pro
The earlier procurement study considers a desk-mounted leader–follower pair, camera, and existing host computer. It remains a reference for later research, not an active purchase plan; no hardware has been purchased for this prototype.
Much later / Preserved research concepts
A robot that works. A place to return.
Mobile assistance and a physical home for robots remain early, unscheduled concepts. We preserve these studies while focusing on first-person datasets; hardware compatibility, operation, and performance still need separate validation.
Mobile cleaning + yard assistance
Explore a smaller mobile robot with an arm for selected cleaning and property tasks, guided by activity context, permitted commands, and clear operating limits.
Review the dimensioned drawing (PDF)AI shed / Robot home
A planned property-based dock with charging and tool storage for a compatible humanoid or smaller mobile arm. The concept gives the robot a place to return, prepare, and wait between tasks.
Review the dimensioned drawing (PDF)Download editable CAD package · Rev A
Includes Blender and OpenSCAD source, full-scale and 1:10 concept STL files, dimensioned drawings, and recorded parameters.
Revision A engineering concepts establish packaging and access. Mounting, structural loads, electrical and charging design, weather resistance, and powered operation remain unvalidated.
Later research / Preserved design study
First-person datasets are our current focus. The wearable backpack and other robotics concepts remain much later research; earlier design studies are preserved below.
A Matha Labs product concept
Robotic Backpack
Two hands. More possibilities.
A wearable backpack with four robotic arms. We’re exploring AI that sees what you’re doing, understands the environment around you, and coordinates extra hands to help with the task at hand.
In development · Working concept
Explore this idea with usWearable robotics / 01
01
See what you see
Observe the wearer’s actions, nearby objects, and the workspace to build a shared picture of the task.
02
Understand the next step
Use the activity and surrounding context to identify where an extra hand could help.
03
Lend four extra hands
Coordinate four backpack-mounted arms to hold, position, and bring objects within reach while you work.
Lab interests
From today’s focus to later research.
01
First-person activity datasets
Start with ten datasets about everyday activities such as eating, sleeping, cooking, and going to the gym. Combine documented sources, clearly identified synthetic examples, and planned opt-in capture.
02
Activity understanding
How can egocentric models combine visual and audio cues to understand what a person is doing, what matters in the scene, and when help would be useful?
03
Spatial interfaces
How can people express intent, guide local actions, and understand what Cloud AI is doing on their behalf through glasses and spatial interfaces?
04
Much later: physical AI & robotics
Preserve the desktop arm, mobile robot, and wearable backpack concepts for later research. Human demonstrations could eventually inform physical assistance, after the dataset foundations and separate hardware validation.
Next step
Discuss an AI challenge.
Start with a free 15-minute introductory call with Revanth. We’ll discuss your training data needs or AI challenge and how Matha Labs can help.