Two-Robot Automated Scanning System

• The paper demonstrates significant gains in scanning accuracy and efficiency through coordinated motion planning and precise workspace segmentation.
• It employs a dual-robot architecture where one robot carries a high-resolution scanner while the other manipulates the artefact for full-surface exposure.
• Quantitative metrics like Chamfer Distance and F-score validate the system's superior 3D reconstruction fidelity and reduced batch scan times over manual workflows.

An automated two-robot scanning system is a coordinated framework in which two robotic manipulators perform synchronized geometric data acquisition across a physical object or region. These systems are used for high-fidelity 3D digitization, industrial inspection, laboratory characterization, medical imaging, and heritage conservation. System performance is governed by principles of coordinated motion planning, robust calibration, occlusion minimization, efficient workspace segmentation, trajectory optimization, and repeatable capture protocols. Recent research demonstrates significant gains in geometric accuracy, scanning efficiency, and reduction of expert labor over semi-automatic or manual workflows (Ahmad et al., 6 Oct 2025).

1. System Architecture

The canonical two-robot scanning system comprises a scanner-equipped manipulator and an auxiliary or tray-handling manipulator. The scanner robot mounts a high-resolution 3D scanning device (e.g., Artec3D structured-light scanner) and executes systematic motion sequences about the digitized object. The auxiliary robot manipulates the artefact’s support structure—a transparent plexiglass tray in cultural heritage contexts—enabling global object reorientation and full-surface exposure.

A sphere-centered parameterization of the workspace underpins motion planning. The scanner’s locus is discretized as a grid of spherical coordinates: polar angle (θ), azimuthal angle (φ), and a fixed radial distance (r), typically between 55 cm and 65 cm for optimal measurement conditions. Segmentation divides the sphere into upper, middle, and lower scan regions, each associated with a distinct set of robot poses and tray reorientations. The tray-handling robot executes a sequence of predefined poses (e.g., $T_{up} = \{ (R_i, t_{up}) \}$), which exposes occluded object surfaces for sequential scanning while ensuring system stability and minimizing vibration.

This approach dispenses with fixed rotary tables and linear actuators, offering flexibility in handling artefacts of heterogeneous geometry, fragility, and value (Ahmad et al., 6 Oct 2025). The robots are synchronized via a finite state machine (FSM) that manages home positioning, segmented scanning, coordinated reorientation, and tray return.

2. Optimized Motion Planning

Motion planning is driven by a decomposition of the scanning space and explicit discretization in spherical coordinates. Waypoint generation follows

$$S = \{ (\theta_i, \phi_j, r) \mid \theta_i = i \cdot \Delta\theta, \phi_j = j \cdot \Delta\phi, i \in [1, c], j \in [1, w] \}$$

with angular resolutions

$$\Delta\theta = \frac{2\pi}{3c}; \qquad \Delta\phi = \frac{2\pi}{w}$$

where $c$ and $w$ are the polar and azimuthal division counts, respectively. The scanning sphere is further segmented for kinematic feasibility: upper ($\theta\in[\frac{\pi}{6},\frac{\pi}{2}]$), middle, and lower ($\theta\in[\frac{\pi}{2},\frac{5\pi}{6}]$) regions.

Inverse kinematics (IK) solutions are computed for each waypoint, subject to joint and workspace constraints of both robots. Tray poses are defined using rotation matrices $R_i \in SO(3)$ and fixed translations. This coordinated scheduling ensures collision-free access to all object surfaces and supports multidirectional data capture without operator intervention.

The FSM manages synchronization, activating sequential tasks such as home position acquisition, scan segment execution, tray reorientation, and final object return. This guarantees both coverage completeness and temporal coordination.

3. Surface Coverage and Occlusion Minimization

The core objective of two-robot scanning is comprehensive surface coverage with minimal occlusions and artifacts.

A transparent plexiglass tray handled by the auxiliary robot mitigates occlusion effects otherwise induced by opaque supports, rotary stages, or manual repositioning. Coordinated reorientations enable exposure of surfaces shielded in prior segments. The scanner-equipped robot’s spherical trajectory is sequenced to maximize direct line-of-sight, while auxiliary robot poses are chosen to minimize artefact motion and vibrational disturbances, leveraging vibration-damping grip materials.

Surface coverage completeness is validated by comparing reconstructed point clouds against high-precision references using bidirectional distance metrics (Chamfer Distance, F-score). Capture reliability is ensured through real-time integration with scanning device SDKs; e.g., real-time focus checks and automated retries yield capture success rates near 99% (Ahmad et al., 6 Oct 2025).

4. Quantitative Performance Metrics

Scanning quality and efficiency are assessed using:

Metric	Purpose	Formula (Paper)
Chamfer Distance	Measures geometric reconstruction error	$$CD(X,Y) = \frac{1}{2|X|}\sum_{x \in X}\min_{y \in Y}||x-y||^2 + \frac{1}{2|Y|}\sum_{y \in Y}\min_{x \in X}||x-y||^2$$
F-score@d	Harmonic mean of precision and recall	$F = 2 \cdot \frac{P \cdot R}{P + R}$; $P$ and $R$ defined on spatial thresholds

Reported experiments indicate lower Chamfer Distance and higher F-score compared to baseline pipelines, indicating higher-fidelity reconstruction and spatial completeness. Batch scan times are reduced (∼9 min per batch) with operator intervention minimized by almost a factor of three over manual workflows. This demonstrates the direct impact of two-robot coordination and optimized spatial parameterization (Ahmad et al., 6 Oct 2025).

5. Technical Challenges and Mitigations

Key challenges include:

• Artefact Reorientation and Occlusions: Solved by using a transparent tray and precise gripper mechanisms.
• Trajectory Optimization Within Kinematic Constraints: Resolved by segmenting workspace and precomputing feasible waypoints and tray poses.
• Robot Synchronization: Managed via a finite state machine that sequences and locks robot actions.
• Stability and Vibration: Addressed through rubberized grip materials and careful tray manipulation.
• Capture Reliability: Real-time focus and scan status feedback with automated retry mechanisms yields high success rates.

The development demonstrates how two-robot systems can surpass manual or semi-automatic pipelines, specifically in scenarios where artefact fragility, geometric complexity, and occlusion avoidance are critical.

6. Applications and Broader Implications

The principal application is high-fidelity digitization of cultural heritage artefacts for long-term preservation, documentation, and dissemination. Automated two-robot scanning is also applicable to industrial quality control, non-destructive testing, reverse engineering, and medical imaging.

Automating expert-intensive workflows reduces dependency on specialized labor, making comprehensive digitization feasible in resource-limited settings. The system design and evaluation framework enable future research in adaptive perception, real-time feedback, and even vision-language enhanced automation. The modular architecture supports easy adaptation to new scanning devices, artefact types, and domain requirements.

7. Research Significance and Future Directions

The automated two-robot scanning system realizes substantial advances in geometric accuracy, efficiency, and workflow automation (Ahmad et al., 6 Oct 2025). Quantitative gains over state-of-the-art methods validate the importance of coordinated manipulation and explicit workspace segmentation.

Extension to more sophisticated sensor modalities, integration with active scene understanding, and progression toward multi-robot cooperative frameworks are plausible directions. A plausible implication is the democratization of cultural heritage digitization and transfer of best practices to new domains where precision, coverage, and automation are essential.

In summary, two-robot scanning systems combine optimized motion planning, robust synchronization, and advanced occlusion mitigation to deliver high-quality, efficient digitization pipelines for complex artefacts and environments.