Job description

execute manual data annotation and QA review: Perception annotation: video, images, machine sensor data that describe objects, 3D, trajectory, etc VLA annotation: task, action, intent, chain of causation, outcomes Build the brain behind the world’s heaviest autonomous machines. We are seeking a highly detail-oriented Data Annotator & QA Reviewer to join our Autonomy & Robotics team and directly shape the future of mining operations. In this role, you won't just label images—you’ll solve complex 3D spatial challenges, fuse multi-sensor telemetry (LiDAR, radar, CAN bus), and map the decision-making logic powering massive haul trucks, excavators, and drills in extreme environments. If you thrive at the intersection of robotics, spatial perception, and high-precision AI data, join us in building the ground-truth foundation for next-generation Vision-Language-Action (VLA) models. Required Skills - Data Annotation Labelling QA for computer vision 3D spatial data LiDAR labelling platforms Job Duties - Key Responsibilities 1. Mining Perception & Spatial Sensor Annotation 3D Spatial Bounding & Segmentation: Annotate site entities (haul roads, berms, rock piles, ore benches, light vehicles, personnel, and machinery) across 2D video feeds and 3D LiDAR/radar point clouds. Heavy Equipment Trajectory Tracking: Map precise motion paths, articulation angles, bucket/blade orientations, and velocity vectors for heavy vehicles operating in constrained mining environments. Sensor Fusion Alignment: Cross-reference visual camera telemetry with heavy vehicle sensors (GPS/GNSS, IMU, CAN bus torque/hydraulic pressure, payload sensors) to maintain temporal and spatial alignment. 2. VLA & Heavy Operational Behavior Mapping Mining Task & Action Decomposition: Segment complex operational workflows into granular actions (e.g., Bench Approach $\rightarrow$ Spotting $\rightarrow$ Bucket Dig Cycle $\rightarrow$ Swing $\rightarrow$ Hopper Dump $\rightarrow$ Haul Cycle). Intent Identification: Identify and label machine and operator intent behind steering shifts, speed adaptations, and bucket/blade maneuvers (e.g., Yielding to light vehicle, Negotiating steep grade, Slippage recovery). Chain of Causation Modeling: Annotate environmental triggers and causal relationships specific to mining conditions (e.g., High dust reduced visibility $\rightarrow$ Speed reduced; Oversized boulder detected in pit $\rightarrow$ Trajectory rerouted). Outcome Verification: Tag expected vs. actual site outcomes (e.g., Full Bucket Load Achieved, Tire Torque Slip, Berm Encroachment, Dumping Clearance Succeeded). 3. Quality Assurance (QA) & Audit Conduct rigorous QA audits on labeled mining datasets to enforce strict precision standards across edge cases (extreme dust, mud, nighttime/glare lighting, subterranean conditions). Audit temporal consistency in vehicle trajectory sequences and ensure correct semantic labeling of mining-specific hazards and terrain features. Provide structured feedback to internal annotators and external data partners, updating labeling schema guidelines as mining edge cases emerge. Job Requirements - Qualifications & Requirements Required Experience 1+ years of professional experience in data annotation, labeling, or QA for computer vision, robotics, or autonomous systems. Experience handling 3D spatial data (LiDAR point clouds, depth maps, spatial trajectories, multi-camera feeds). Working knowledge of standard labeling platforms (CVAT, Labelbox, Scale AI, Supervisely, V7, Encord, etc.). Ability to break down complex heavy-machinery interactions into structured sequence flows: Task $\rightarrow$ Action $\rightarrow$ Intent $\rightarrow$ Causation $\rightarrow$ Outcome. Key Technical & Soft Skills Domain Literacy: Familiarity with mining operations, pit safety terminology, and heavy equipment mechanics (haulers, excavators, loaders). Spatial Perception: Strong 3D spatial visualization skills (understanding vehicle yaw/pitch/roll, bucket kinematics, and 3D point cloud depths). Attention to Detail: Meticulous approach to labeling tight bounding boxes and subtle terrain/hazard changes in poor visibility conditions. Technical Aptitude: Comfortable working with geospatial formats, sensor logs, and structured metadata formats (JSON/XML). Desired Skills & Experience - Nice to have Background in Mining Engineering, Geotechnical Engineering, Robotics, Autonomous Vehicles, or Agricultural/Industrial Autonomy. Experience with autonomous haulage systems (AHS), telemetry logs, or embodied VLA models for industrial robotics.

Requirements

Must-haves

execute manual data annotation and QA review: Perception annotation: video, images, machine sensor data that describe objects, 3D, trajectory, etc VLA annotation: task, action, intent, chain of causation, outcomes
Required Skills - Data Annotation
QA for computer vision
3D spatial data
labelling platforms
Job Duties - Key Responsibilities
Mining Perception & Spatial Sensor Annotation
3D Spatial Bounding & Segmentation: Annotate site entities (haul roads, berms, rock piles, ore benches, light vehicles, personnel, and machinery) across 2D video feeds and 3D LiDAR/radar point clouds.
Heavy Equipment Trajectory Tracking: Map precise motion paths, articulation angles, bucket/blade orientations, and velocity vectors for heavy vehicles operating in constrained mining environments.
Sensor Fusion Alignment: Cross-reference visual camera telemetry with heavy vehicle sensors (GPS/GNSS, IMU, CAN bus torque/hydraulic pressure, payload sensors) to maintain temporal and spatial alignment.
VLA & Heavy Operational Behavior Mapping
Intent Identification: Identify and label machine and operator intent behind steering shifts, speed adaptations, and bucket/blade maneuvers (e.g., Yielding to light vehicle, Negotiating steep grade, Slippage recovery).
Outcome Verification: Tag expected vs. actual site outcomes (e.g., Full Bucket Load Achieved, Tire Torque Slip, Berm Encroachment, Dumping Clearance Succeeded).
Quality Assurance (QA) & Audit
Conduct rigorous QA audits on labeled mining datasets to enforce strict precision standards across edge cases (extreme dust, mud, nighttime/glare lighting, subterranean conditions).
Audit temporal consistency in vehicle trajectory sequences and ensure correct semantic labeling of mining-specific hazards and terrain features.
Provide structured feedback to internal annotators and external data partners, updating labeling schema guidelines as mining edge cases emerge.
Job Requirements - Qualifications & Requirements
1+ years of professional experience in data annotation, labeling, or QA for computer vision, robotics, or autonomous systems.
Experience handling 3D spatial data (LiDAR point clouds, depth maps, spatial trajectories, multi-camera feeds).
Working knowledge of standard labeling platforms (CVAT, Labelbox, Scale AI, Supervisely, V7, Encord, etc.).
Ability to break down complex heavy-machinery interactions into structured sequence flows: Task $\rightarrow$ Action $\rightarrow$ Intent $\rightarrow$ Causation $\rightarrow$ Outcome.
Key Technical & Soft Skills

Nice-to-haves

Desired Skills & Experience - Nice to have

Application prep

Use the role’s language and application details while tailoring your materials.

Keywords to mirror

Platform: jobdiva
Effort: medium

Benefits

Benefits detected in the listing.

Vision