Dragonflies are agile and efficient flyers that use two pairs of wings for demonstrating exquisite aerial maneuvers. Compared to two-winged insects such as bees or flies, dragonflies leverage forewing and hindwing interactions for achieving higher efficiency and net lift. Here we develop the first at-scale dragonfly-like robot and investigate the influence of flapping-wing kinematics on net lift force production. Our 317 mg robot is driven by two independent dielectric elastomer actuators that flap four wings at 350 Hz. We extract the robot flapping-wing kinematics using a high-speed camera, and further measure the robot lift forces at different operating frequencies, voltage amplitudes, and phases between the forewings and hindwings. Our robot achieves a maximum lift-to-weight ratio of 1.49, and its net lift force increases by 19% when the forewings and hindwings flap in-phase compared to out-of-phase flapping. These at-scale experiments demonstrate that forewing–hindwing interaction can significantly influence lift force production and aerodynamic efficiency of flapping-wing robots with passive wing pitch designs. Our results could further enable future experiments to achieve feedback-controlled flights.
To close the gap between conceptual ideas and applied solutions, soft roboticists must build physical hardware. How that hardware is built meaningfully impacts its performance, as soft robots are highly sensitive to manufacturing. When fabricating components, researchers must select a high‐level process and numerous implementation details but currently lack both characterization data to make informed choices and a strategy for evaluating process tradeoffs. This paper addresses both gaps for molding soft elastomeric components, comparing injection molding (IM) and hand casting (HC) across molds from three additive manufacturing technologies: fused filament fabrication (FFF), stereolithography (SLA), and PolyJet (PJ). Using a representative inflatable geometry, actuators, molds, and test coupons are characterized with cycle testing, optical microscopy, profilometry, and 3D imaging. Lab‐scale, soft robotics‐specific criteria for evaluating fabrication strategies are defined and applied. Results demonstrate no universally best strategy exists; appropriate choices are context‐dependent. For example, SLA and PJ molds produced inflatable actuators lasting 2.5–3× longer than FFF, but at orders‐of‐magnitude greater cost and longer design‐to‐test times. Similarly, IM yielded better‐aligned internal cores than HC but with higher consumable costs and more hands‐on time. Synthesizing these criteria with characterization data and representative fabrication scenarios provides concrete process selection recommendations for matching fabrication with application.