Scalable Low-Density Distributed Manipulation Using an Interconnected Actuator Array

Bailey Dacre; Rodrigo Moreno; Jørn Lambertsen; Kasper Stoy; Andrés Faíña

cs.ROFeb 23, 2026

Scalable Low-Density Distributed Manipulation Using an Interconnected Actuator Array

Authors: Bailey Dacre, Rodrigo Moreno, Jørn Lambertsen, Kasper Stoy, Andrés Faíña

Abstract

Distributed Manipulator Systems, composed of arrays of robotic actuators necessitate dense actuator arrays to effectively manipulate small objects. This paper presents a system composed of modular 3-DoF robotic tiles interconnected by a compliant surface layer, forming a continuous, controllable manipulation surface. The compliant layer permits increased actuator spacing without compromising object manipulation capabilities, significantly reducing actuator density while maintaining robust control, even for smaller objects. We characterize the coupled workspace of the array and develop a manipulation strategy capable of translating objects to arbitrary positions within an N X N array. The approach is validated experimentally using a minimal 2 X 2 prototype, demonstrating the successful manipulation of objects with varied shapes and sizes.

Figures & tables

Fig. 1 : A $2\times 2$ actuator array inter-connected by a flexible material to form a continuos manipulation surface

Fig. 2 : Array hardware: (a) Robotic tile module comprising three origami-inspired legs, each actuated by a stepper motor and connected to a central end-effector with attachment points for the flexible surface. (b) Detailed view of a single leg assembly. (c) Visualization of the compliant surface layer showing regions of exposed rubber and bonded PVC film.

Fig. 3 : Illustration of four modules modules in an array, connected by connective material. This shows motor angles

\theta_{i}

, which determine tile pose

(\delta^{i},\phi^{i},r^{i})

. Modules are at an inter-module distance,

D

, and are connected by a material of length

L

, with distances

\alpha

and

\beta

indicated.

Fig. 4 : Visualization of the workspace of a single tile in red, and a radially symmetric subsection of that workspace in blue. Points show position of end-effector centre.

Fig. 5 : Minimum material length to achieve all positions in single tile workspace when connected to a horizontally/ vertically adjacent tile ( $\alpha_{max}$ ) and diagonally adjacent tile ( $\beta_{max}$ ) for different inter-tile distances (D).

Fig. 6 : Effect of inter-tile material length on tile pose. One tile remains fixed while the other is positioned at $D=$261\text{\,}\mathrm{mm}$$ with $\delta$ directed away. (a) Maximum reachable $\phi$ and (b) maximum $\gamma$ for different $r$ values. The grey line indicates the material length used in the physical prototype. — Fig. 6 : Effect of inter-tile material length on tile pose. One tile remains fixed while the other is positioned at $D=$ 261\text{,}\mathrm{mm}$$ with $\delta$ directed away. (a) Maximum reachable $\phi$ and (b) maximum $\gamma$ for different $r$ values. The grey line indicates the material length used in the physical prototype.

Current Region $\rightarrow$ Next Region	Tile NW	Tile NE	Tile SW	Tile SE
TILE $\rightarrow$ INTER_TILE	$(0,\frac{5\pi}{36},90)$	(0, 0, 90)	(0, 0, 90)	(0, 0, 90)
INTER_TILE $\rightarrow$ TILE	$(0,0,10)$	$(0,\frac{5\pi}{36},90)$	(0, 0, 90)	(0, 0, 90)
INTER_TILE $\rightarrow$ CENTRE	$(\frac{3\pi}{2},\frac{5\pi}{36},90)$	$(\frac{3\pi}{2},\frac{5\pi}{36},90)$	$(\frac{\pi}{2},\frac{5\pi}{36},90)$	$(\frac{\pi}{2},\frac{5\pi}{36},90)$
CENTRE $\rightarrow$ INTER_TILE	$(0,0,10)$	$(\pi,0,10)$	$(\frac{5\pi}{4},\frac{\pi}{12},90)$	$(\frac{7\pi}{4},\frac{\pi}{12},90)$

TABLE I : Canonical states determining tile pose from current and target regions, listed in the form

(\delta,\phi,r)

. All other transitions are obtained by

R(\frac{\pi}{2}k)

rotation symmetry,

k\in\{0,1,2,3\}

. For example,

k=0

corresponds to the tile located in the north-west position of the array. Angles are in radians, heights in millimetres.

Fig. 8 : Trajectory of puck performing three cyclic translation around the array. Intermediary targets highlighted.

Fig. 9 : trajectory of sphere performing cyclic translation between inter-tile regions. No distinct targets positions, as here targets are just region transitions.

Fig. 10 : Trajectories of cube moving towards a an arbitrary target (150.5mm,-110.5mm) positioned on a tile, translating from multiple starting positions.

Fig. 11 : Trajectories of the tetrahedron moving towards a an arbitrary target (-155.5mm, 115.5mm) positioned on a tile, translating from multiple starting positions.

Robotic Manipulation

Robotic manipulation studies how robots perceive, plan, and control physical interactions to grasp, move, assemble, and use objects.

See all 1,685 papers

Making Waves: A Membrane-Coupled Delta Array for Manipulating Objects Below the Actuator Spacing
Sep 30, 2026Bailey Dacre, Andrés Faíña, Oliver Kroemer +1Robotic Control Robotic Manipulation
Distributed Dexterous Manipulation with Spatially Conditioned Multi-Agent Transformers
Sep 7, 2026Sarvesh PatilRobotic Manipulation Multi-Robot Coordination
Carnegie Mellon University, The Robotics Institute Pittsburgh, PA, United States
One Print, Many Moves: Monolithic Origami-inspired Folding Actuator for Composable Soft Multi-DoF Systems
Sep 1, 2026Jaehyung Jang, Zhenish Zhakypov, Jasmin Elena Palmer +3Modular Robotics Robotics
IRiS Lab, Department of Civil and Environmental Engineering, Korea Advanced Institute of Science and Technology, Daejeon, Korea. · CHARM Lab, Department of Mechanical Engineering, Stanford University, Stanford, CA, USA.

Sep 30, 2026J/K move · Enter open · S save

Making Waves: A Membrane-Coupled Delta Array for Manipulating Objects Below the Actuator Spacing

Bailey Dacre, Andrés Faíña, Oliver Kroemer, Zeynep Temel

Robotic Control Robotic Manipulation

Open paper AskPDF arXiv

Distributed manipulator systems manipulate objects through the coordinated motion of many actuators. However, an object must be supported by several actuators at once, so the centre-to-centre actuator spacing imposes a hard lower bound on manipulable object size. We remove this bound by coupling the end-effectors of an 8 x 8 array of three degrees-of-freedom delta robots with a stretchable fabric, turning 64 discrete contacts into a continuous surface capable of manipulating objects smaller than the actuator spacing. Viewing the array as a displacement field over that surface, we investigate local quasi-static and cyclic fields as manipulation primitives. These primitives can be applied globally across the array or locally confined around each tracked object to independently manipulate several objects in parallel. We then train a policy acting on low-order discrete cosine transform coefficients: at equal action dimension, commanding a nineteen-delta neighbourhood halves the placement error of commanding the whole array. The policy transfers to hardware without adaptation at 76% success. The platform manipulates objects from 15mm-90mm, a six-fold range spanning both sides of the actuator spacing, 43.3mm, on a single surface.

Abstract

Robotic Manipulation

Explore similar work

Making Waves: A Membrane-Coupled Delta Array for Manipulating Objects Below the Actuator Spacing

Distributed Dexterous Manipulation with Spatially Conditioned Multi-Agent Transformers

One Print, Many Moves: Monolithic Origami-inspired Folding Actuator for Composable Soft Multi-DoF Systems