Seeking Partners for Large-Scale Egocentric Data Collection for Ai & Robotics Training Models

We are expanding our egocentric (first-person) video and sensor data collection operations into Latin America and are seeking staffing and data-collection partners. This large-scale program aims to gather data for training advanced embodied-AI and robotics models. Project Scope: Total expected volume: Up to 1 million QC-approved hours. Data Type: Egocentric footage (first-person video) captured with head- or forehead-mounted devices. The camera must be angled downward to ensure both palms and all fingers remain visible while manipulating objects. Sensor Data: Synchronized IMU/pose sensor data will also be collected. Privacy and Quality Guidelines (Non-Negotiable): Recording will only occur during active task performance. No personal faces or identifiable ID information should be in frame. Footage will never be shared publicly. Documented consent must be obtained at every collection site. Payment is based on QC-approved hours, emphasizing that quality is integrated into the process from the outset. Environments for Data Collection: 1. Residential Environments (equal split across four task families): Laundry tasks, Kitchen tidy-up (excluding cooking), Organizing a single room, Other general household tasks. 2. Commercial Environments (spanning various real workplaces): Cloud kitchens, Warehouses, Hospitality & housekeeping, Confectioneries, Packaging, FMCG repacking & white-labelling, Waste management, Skilled trade, Workshop, Field-services light installation, Other approved categories. Technical Requirements for Capture Devices: Accepted phones: Recent high-end smartphones such as iPhone 12 and newer, Google Pixel 6 and newer, and Samsung Galaxy S21 and newer. Camera mounting: Devices must be mounted at forehead/eye level, angled down to ensure full hand visibility. We are looking for experienced partners capable of managing and executing data collection with strict adherence to quality and privacy protocols.