Enabling Tidying Tasks in Robotic Vacuum Cleaners via a Geometry-Induced Self-Locking Deployable Linkage

Eunseok Song; Jinwoong Kim; Sohee Lim; Taehyeong Kim; Yongkook Kim; Hyeontaek Lim; Hyungmin Son; Seungyong Hyung

cs.ROOct 14, 2026

Enabling Tidying Tasks in Robotic Vacuum Cleaners via a Geometry-Induced Self-Locking Deployable Linkage

Authors: Eunseok Song, Jinwoong Kim, Sohee Lim, Taehyeong Kim, Yongkook Kim, Hyeontaek Lim, Hyungmin Son, Seungyong Hyung

Organizations: Samsung Research, Samsung Electronics, 56, Seongchon-gil, Seocho-gu, Seoul, South Korea.

Abstract

Robotic vacuum cleaners must extend their functionality within tight limits on size, cost, and durability. Deployable linkages with few actuators meet these limits, but their backdrivability makes it difficult to hold a load-bearing posture without continuous motor torque. We introduce a two-stage deployable linkage that locks through its own geometry. At full deployment, an internal three-link subchain approaches a collinear, near-singular configuration in which the input-output transmission ratio tends to zero; plate loads therefore leave only a small residual back-driving torque, which drive-train friction and a lightweight magnetic detent absorb. A single actuator performs deployment and locking, a passive chassis-mounted trigger releases the lock for retraction, and a second actuator adjusts only the deployment height. We integrate the mechanism into a commercial robotic vacuum cleaner and validate it through a 100-cycle repeatability test, a quasi-static load test in which the locked plate held 50 N with the motor de-energized, and 140 pushing-type tidying trials on seven household objects against a commercial arm-equipped vacuum robot. We derive the analytical holding condition and discuss its limits as an intermediate design point between passive compliance and articulated manipulation.

Figures & tables

Fig. 1 : A commercial robotic vacuum cleaner equipped with the proposed geometry-induced self-locking deployable linkage. A single deployment actuator transitions the two-stage plates between a backdrivable deployment motion and a near-singular, load-supporting configuration while preserving the original stowed form factor.

Panel of Figure 2 — Fig. 2 : Operating states of the six-link deployable mechanism. $\theta_{\mathrm{in}}$ is the rotation of the driven joint $j_{1}$ from the stowed configuration and $\theta_{\mathrm{out}}$ the rotation of $l_{2}$ relative to the chassis. (a) Backdrivable deployment : far from alignment the transmission ratio $J$ is $O(1)$ ; the plates first rotate together about $j_{1}$ with the internal loop folded, then $l_{2}$ opens relative to $l_{1}$ . (b) Near-singular rigidification : the internal subchain $l_{4}$ – $l_{6}$ – $l_{2}$ approaches collinearity and $J\to 0$ , so that plate loads are carried axially along the aligned members. (c) Body-triggered release : a chassis protrusion perturbs the aligned joint during retraction, restoring $J$ and allowing folding.

Fig. 3 : Robotic vacuum with the deployable rigidifying mechanism. Transparent silhouettes show the unfolding and rigidifying sequence as the two-stage plates engage an object; the chassis-mounted release trigger is indicated.

Panel of Figure 4 — Fig. 4 : 100-cycle repeatability test. (a) Marker trajectories at the distal ends of the second-stage links in overhead-camera image coordinates; the black cross is the kinematic origin (marker midpoint in the stowed state). (b) Opening angle between the rays from the origin to the two markers versus normalized motion phase for all 100 cycles; the inset shows the locked state. The opening angle is a monotonic function of $\theta_{\mathrm{in}}$ and $\theta_{\mathrm{out}}$ and is used because it is directly observable from the markers.

Fig. 5 : Quasi-static in-plane loading of the locked plate. The chassis is clamped to a fixture; a non-elastic cable runs from the plate tip through an inline tensile load cell, in the direction opposite to sweeping (arrow).

Fig. 6 : Representative tidying trials on six of the seven test objects (the wadded tissue is omitted for space). Each trial consists of approach, deployment and locking, a single forward push to the target zone, retraction, and departure.

Object	Category	Proposed	Bumper	Arm
Plastic toy (Lego blocks)	rigid, small	100%	0%	0%
Container (Lock&Lock)	rigid, bulky	100%	0%	0%
Toy car (die-cast)	rigid, small	100%	0%	20%
Towel	deformable, wide	100%	0%	0%
Tissue (wadded)	deformable, light	100%	0%	100%
Indoor slipper	low-profile	100%	0%	50%

TABLE I : Tidying success rate over 20 trials per object for each system. “Bumper” is the unmodified base platform (Samsung VR9700); “Arm” is the Roborock Saros Z70.

Mechanism Design

Mechanism design studies rules for allocating resources and shaping incentives when participants hold private information or act strategically.

See all 142 papers

DAWN: Diffusion-based Alignment and unfastening Workflow Network
Oct 29, 2026Hemant Kumar, Piyush Goenka, Nikhil ChopraRobotic Manipulation Contact-Rich Robotic Manipulation
Department of Mechanical Engineering, University of Maryland, College Park, MD, USA
MiniUGV $_2$ : A Compact UAV-Deployable Tracked Ground Vehicle with Manipulation Capabilities
Mar 1, 2026Durgakant Pushp, Swapnil Kalhapure, Shaekh Mohammad Shithil +1Robotics Mobile Robotics
Luddy School of Informatics, Computing, and Engineering at Indiana University, Bloomington, IN 47408, USA. · S. Kalhapure is an Independent Researcher at the time of work.
Behavioral Cloning for Robotic Connector Assembly: An Empirical Study
Feb 25, 2026Andreas Kernbach, Daniel Bargmann, Werner Kraus +1Robotic Insertion Robotic Manipulation
The authors are with the Fraunhofer Institute for Manufacturing Engineering and Automation IPA, Stuttgart, Germany · Institute of Industrial Manufacturing and Management IFF, University of Stuttgart, Germany

Oct 29, 2026J/K move · Enter open · S save

DAWN: Diffusion-based Alignment and unfastening Workflow Network

Hemant Kumar, Piyush Goenka, Nikhil Chopra

Department of Mechanical Engineering, University of Maryland, College Park, MD, USA

Robotic Manipulation Contact-Rich Robotic Manipulation Long-Horizon Robotic Manipulation Force-Controlled Robotic Manipulation

Open paper AskPDF arXiv

We present DAWN (Diffusion-based Alignment and unfastening Workflow Network), a hierarchical robotic framework designed to perform real-world, unfastening tasks through multi-modal sensory integration. Autonomous robotic unfastening is crucial for improving recycling processes, enhancing safety, and reducing operational costs; yet, it remains predominantly manual due to inherent challenges in precise screwdriver engagement with fasteners, detailed force execution, and robustness to environmental variability. To address the complexity of learning long-horizon, contact-rich manipulation tasks, we subdivide the unfastening task procedure into two sequential stages: (1) alignment & insertion, and (2) unfastening. DAWN consists of three key components: AlignNet, a diffusion-based policy ensuring precise screwdriver tip engagement; LatchNet, a multi-modal classifier reliably detecting successful stage transitions; and UnfastNet, which achieves stable, force-controlled screw removal. Experimental validation on a Franka Emika robot demonstrates that DAWN achieves alignment & insertion success rates comparable to trained human operators, robustly transitions between stages, and attains an impressive success rate in unfastening across various unseen scenarios. Ablation studies further elucidate critical input modalities and effective training strategies, highlighting DAWN's scalability, robustness, and data efficiency.

Abstract

Mechanism Design

Explore similar work

DAWN: Diffusion-based Alignment and unfastening Workflow Network

MiniUGV2_22​: A Compact UAV-Deployable Tracked Ground Vehicle with Manipulation Capabilities

Behavioral Cloning for Robotic Connector Assembly: An Empirical Study

MiniUGV $_2$ : A Compact UAV-Deployable Tracked Ground Vehicle with Manipulation Capabilities