About

Robert J. Wood is the Harry Lewis and Marlyn McGrath Professor of Engineering and Applied Sciences at Harvard University. He serves as the Director of Graduate Studies at the Harvard John A. Paulson School of Engineering and Applied Sciences. His primary teaching areas include Materials Science and Mechanical Engineering. His research encompasses a broad range of fields including applied mathematics, artificial intelligence, machine learning, modeling physical and biological phenomena, applied physics, soft matter, bioengineering, bioinspired robotics and computing, biomechanics and motor control, electrical engineering, robotics and control, and materials science and mechanical engineering. His work involves developing innovative robotic systems, studying biological and physical systems, and advancing technologies in bioengineering and robotics.

Research topics

• Computer Science
• Artificial Intelligence
• Engineering
• Embedded system
• Materials science
• Physics
• Mechanical engineering
• Human–computer interaction
• Computer vision
• Electrical engineering
• Simulation
• Electronic engineering
• Composite material
• Nanotechnology
• Control engineering
• Optics
• Classical mechanics
• Biological system
• Biology
• Mechanics
• Distributed computing
• Theoretical computer science
• Aerospace engineering

Selected publications

• Enabling scalable manufacturing of a microrobotic end-effector for surgical laser steering

  Journal of Micromechanics and Microengineering · 2026-04-01

  articleSenior author

  Abstract Robust millimeter-scale mechanisms have the potential to enable a new generation of end effectors for minimally invasive procedures, aiding surgeons in performing complex tasks with improved precision, repeatability, and dexterity. Despite ongoing advances in fabrication methods, the manufacturing and actuation of complex articulated mechanisms at the millimeter scale involves significant challenges, particularly for integration and assembly. Here, we present the third generation of a complex optoelectromechanical device capable of precisely steering a fiber-delivered surgical laser along two axes of motion. The electromagnetically-driven end effector has been fully redesigned with an emphasis on manufacturability and enabling future high-volume production. The device is 3.8 mm in diameter (4 mm with housing), 10 mm in length, and features ±15 degrees range of motion in both axes with a bandwidth of <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML" overflow="scroll"> <mml:mrow> <mml:mo>&gt;</mml:mo> </mml:mrow> </mml:math> 230 Hz. We anticipate this type of end effector to have applications in tissue ablation and excision in endoscopic procedures throughout the gastrointestinal tract.

  PublisherDOI

• Motion-Uncertainty-Aware Next-Best-View Planning for Moving Object Reconstruction

  arXiv (Cornell University) · 2026-05-17

  preprintOpen access

  Active 3D reconstruction of moving objects requires selecting informative viewpoints while accounting for object motion uncertainty during the decision-to-execution delay. Existing methods address only parts of this problem: next-best-view (NBV) planners for object reconstruction typically optimize surface coverage but assume static objects, while motion-aware active perception for moving targets accounts for target motion but prioritizes tracking or visibility over reconstruction coverage. This work presents a motion-uncertainty-aware NBV framework for reconstructing an unknown rigid object undergoing planar motion, using noisy planar position measurements of the object and depth observations from a mobile robot. The key idea is to evaluate each candidate viewpoint by its expected observation quality over plausible future object states induced by motion and measurement uncertainty, rather than at a single predicted object pose. To obtain this predictive belief, a fixed-lag Gaussian Process smoother estimates and predicts the object state from noisy position measurements. The resulting belief is used to generate candidate viewpoints around the predicted object location, filter them by reachability, and estimate their expected coverage-driven scores. Simulation and real-world experiments demonstrate improved reconstruction completeness over non-predictive NBV and prediction-only tracking methods, bridging coverage-driven active reconstruction and prediction-driven tracking.

  PublisherDOI

• PneuGrasp: Computational Design of Soft Robots for State-Specific Grasping

  Journal of Mechanical Design · 2026-04-27

  article

  Abstract Soft grippers, used in applications such as food handling and assistive devices, leverage multiple soft fluidic actuators (SFAs) for safe and compliant grasping. Designing SFAs is challenging because they must satisfy multiple functional requirements while operating outside the principles of rigid machine design, as they undergo large deformations and exhibit material nonlinearity. Because fabricating numerous design candidates is costly, computational tools have emerged to expedite the search for optimal designs. However, existing computational tools do not focus on SFA design optimization for state-specific grasping, where actuators are optimized for a particular deformation dictated by the intended use case. Moreover, many existing tools support a limited range of performance metrics and optimization modes. Here, we present PneuGrasp, an open-source tool for the design optimization of SFAs according to a user-specified grasping task. The tool can analyze design candidates across multi-functional combinations of seven performance metrics, including the understudied metrics of grasping force, bandwidth, and actuation energy. In addition, PneuGrasp supports three optimization modes that together provide parameter intuition and shorten optimization time. Through a series of examples evaluating over 1,000 design candidates, we demonstrate that PneuGrasp can identify optimized designs that outperform our baseline. For instance, one design achieved a 60 % reduction in maximum strain and a 52 % reduction in actuation volume, while another showed a 405 % decrease in a combined durability–grasping-force performance score. We fabricated and tested over 30 actuators across five distinct designs, demonstrating PneuGrasp's relative prediction capabilities.

  PublisherDOI

• Motion-Uncertainty-Aware Next-Best-View Planning for Moving Object Reconstruction

  ArXiv.org · 2026-05-17

  articleOpen access

  Active 3D reconstruction of moving objects requires selecting informative viewpoints while accounting for object motion uncertainty during the decision-to-execution delay. Existing methods address only parts of this problem: next-best-view (NBV) planners for object reconstruction typically optimize surface coverage but assume static objects, while motion-aware active perception for moving targets accounts for target motion but prioritizes tracking or visibility over reconstruction coverage. This work presents a motion-uncertainty-aware NBV framework for reconstructing an unknown rigid object undergoing planar motion, using noisy planar position measurements of the object and depth observations from a mobile robot. The key idea is to evaluate each candidate viewpoint by its expected observation quality over plausible future object states induced by motion and measurement uncertainty, rather than at a single predicted object pose. To obtain this predictive belief, a fixed-lag Gaussian Process smoother estimates and predicts the object state from noisy position measurements. The resulting belief is used to generate candidate viewpoints around the predicted object location, filter them by reachability, and estimate their expected coverage-driven scores. Simulation and real-world experiments demonstrate improved reconstruction completeness over non-predictive NBV and prediction-only tracking methods, bridging coverage-driven active reconstruction and prediction-driven tracking.

  PublisherOA PDF

• Real-time Remote Tracking and Autonomous Planning for Whale Rendezvous using Robots

  ArXiv.org · 2025-12-05

  preprintOpen access

  We introduce a system for real-time sperm whale rendezvous at sea using an autonomous uncrewed aerial vehicle. Our system employs model-based reinforcement learning that combines in situ sensor data with an empirical whale dive model to guide navigation decisions. Key challenges include (i) real-time acoustic tracking in the presence of multiple whales, (ii) distributed communication and decision-making for robot deployments, and (iii) on-board signal processing and long-range detection from fish-trackers. We evaluate our system by conducting rendezvous with sperm whales at sea in Dominica, performing hardware experiments on land, and running simulations using whale trajectories interpolated from marine biologists' surface observations.

  PublisherOA PDFDOI

• Individual and Collective Behaviors in Soft Robot Worms Inspired by Living Worm Blobs

  2025-05-19 · 4 citations

  article

  California blackworms constitute a recently identified animal system exhibiting unusual collective behaviors, in which dozens to thousands of worms entangle to form a “blob” capable of actions like locomotion as an aggregate. In this paper we describe a system of pneumatic soft robots inspired by the blackworms, intended for the study of collective behaviors enabled and mediated by such physical entanglement. Both the robots and worms have high aspect ratio (<tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$\gtrsim 1: 50$</tex>), intertwine in complex 3D configurations, operate both in air and underwater, and can locomote both individually and as a collective. We demonstrate and characterize locomotion for both individual robots and entangled blobs, explore the tunability of entanglement strength, and compare these to the analogous versions in living worms. The robots provide a testbed for studying mechanisms underlying behaviors observed in worm blobs, as well as serving as a platform for studies of novel collective behaviors based on physical entanglement.

  PublisherDOI

• Compliance for enhanced electroadhesion: Designing kirigami patterns for local conformation on rough surfaces

  Applied Physics Letters · 2025-02-17

  articleSenior author

  We aim to improve the adhesion capabilities of electroadhesive pads on rough surfaces by using geometry-driven compliance to increase effective contact area. We present a kirigami-based approach for enhancing compliance through an exploration of geometric features cut into an adhesive disk. We experimentally test a range of geometries, comparing shear adhesion strength to understand structure–function relationships in our chosen parameter space. Our findings indicate that introducing cuts to form serpentine paths in a disk results in longer effective lengths and enhanced compliance, thus requiring less energy to deform into a rough surface. Leveraging this insight and associated scaling analysis, we conclude that serpentine-like features arranged in a radially symmetric wedge configuration achieve high levels of adhesion, even on rough surfaces, enabling robust adhesion relative to featureless electroadhesive disks.

  PublisherDOI

• Sticking the landing: Insect-inspired strategies for safely landing flapping-wing aerial microrobots

  Science Robotics · 2025-04-16 · 15 citations

  articleSenior authorCorresponding

  For flying insects, the transition from flight to surface locomotion requires effective touchdown maneuvers that allow stable landings on a variety of surfaces. Landing behaviors of insects are diverse, with some using more controlled flight approaches to landing, whereas others dampen collision impacts with parts of their bodies. The landing approaches of real insects inspired our current work, where we present a combined mechanical and control approach to achieving safe and accurate landings for flapping-wing microaerial vehicles. For the mechanical approach to landing, we took inspiration from the legs of the crane fly, designing lossy compliant legs that maximize energy dissipation during surface collisions. We explored three features in the compliant leg design: leg stance, number of joints, and joint placement. For the control approach to landing, the challenge lies in overcoming the aerodynamic ground effect near the surface. Leveraging the compliant leg design during impact, we designed the preimpact behavior, drawing inspiration from insect landing trajectories, to increase landing success. The proposed controlled landing sequence includes an initial acceleration from hovering, followed by deceleration toward the target, ending with a nonzero impact velocity, similar to what is observed in insects. Last, using an insect-scale flapping-wing aerial microrobot platform (Harvard RoboBee), we verified the controlled, safe, and accurate landing on natural terrain.

  PublisherDOI

• Bioinspired surface structures for added shear stabilization in suction discs

  Scientific Reports · 2025-01-06 · 4 citations

  articleOpen accessSenior author

  Many aquatic organisms utilize suction-based organs to adhere to diverse substrates in unpredictable environments. For multiple fish species, these adhesive discs include a softer disc margin consisting of surface structures called papillae, which stabilize and seal on variable substrates. The size, arrangement, and density of these papillae are quite diverse among different species, generating complex disc patterns produced by these structures. Considering papillae arrangements in three fish species, the Northern Clingfish (Gobiesox maeandricus), Tidepool Snailfish (Liparis florae), and Chilean Clingfish (Sicyases sanguineus), we fabricated physical disc models that tested relevant surface pattern parameters under shear loading conditions. Parameters of interest included the area of papillae-like structures, the spacing between adjacent structures (channel spacing), and the percent coverage of elements relative to the total disc area. To create our models, a soft silicone elastomer was added to a stiff circular suction cup, which was then "stamped" using a laser-etched and thermoformed mold base to create the desired surface patterning. Discs were tested using a robotic arm equipped with a force sensor, which sheared them across smooth and rough surfaces at a fixed speed and distance. The arm was also used to vary the initial compression to test performance under both suction-dominant and friction-dominant preloads. For our designs, patterns with smaller papillae-like structures and channel spacing often produced higher peak forces than those with larger features. However, the design that withstood the highest shear load featured an intermediate pad size and channel spacing, potentially highlighting a balance between overall surface area and fluid channeling. Additionally, discs with surface patterns often outperformed the control discs (no pattern) on both smooth and rough surfaces, but performance was highly dependent on preload, with patterned discs exhibiting benefits with the higher "friction-dominant" preloads.

  PublisherOA PDFDOI

• Magnetic Sensing for Proprioception of Rolling Contact Joints

  2025-04-22 · 1 citations

  articleSenior author

  Rolling contact joints are an advantageous building block for soft-rigid hybrid robots due to their out-of-plane compliance, low resistance to bending, and wide range of motion. Typically, proprioception in these cable-driven mechanisms is achieved through torque and position sensors at the servomotor actuator. However, this indirect measurement of joint position can be inaccurate when the system is non-ideal (e.g., the cables are extensible, friction is not negligible, or the linkage encounters an external disturbance). Here, we introduce a magnetic sensor integrated into rolling contact joints to measure joint position and force. We then explore the design space of this magnetic sensor by varying the orientation of the magnets and the stiffness of the magnetoelastomer. A multilayer perceptron is used to relate the joint position and force to the magnetometer readings, providing the flexibility to use this sensor with various joint geometries and sensing modalities. We find that the accuracy of this sensing and modeling approach is coupled with the configuration of the magnetic elements, and that our system can predict the joint angle, twist, and contact forces with errors as low as 1.6%, 0.1%, and 8.8%, respectively. This work describes a method for enabling proprioception in rolling contact joints, and, more broadly, compliant joints for rigid-soft hybrid robots.

  PublisherDOI

Recent grants

• Collaborative Research: A Combustion-Powered, Flapping-Wing Micro Air Vehicle

  NSF · $226k · 2015–2018

• Collaborative Research: RoboBees: A Convergence of Body, Brain and Colony

  NSF · $9.3M · 2009–2014

• NRI-Large: Collaborative Research: Soft Compliant Robotic Augmentation for Human-Robot Teams

  NSF · $520k · 2012–2017

• 2012 Waterman Award

  NSF · $1.0M · 2012–2018

• RI: Medium: Collaborative Research: Novel microLIDAR Design and Sensing Algorithms for Flapping-Wing Micro-Aerial Vehicles

  NSF · $302k · 2015–2019

Frequent coauthors

• Roger H. French

  Case Western Reserve University

  66 shared

• Michael Karpelson

  Harvard University

  56 shared

• Néstor O. Pérez-Arancibia

  49 shared

• Benjamin M. Finio

  Cornell University

  47 shared

• John P. Whitney

  45 shared

• Daniel M. Vogt

  Harvard University

  45 shared

• Conor J. Walsh

  Harvard University

  44 shared

• Pratheev S. Sreetharan

  38 shared

Labs

• Microrobotics LabPI