MIT Advances Soft Robotics with New Research at IEEE RoboSoft 2026
Researchers at the Massachusetts Institute of Technology (MIT) are pushing the boundaries of soft robotics, presenting five papers at the IEEE RoboSoft 2026 conference in Kanazawa, Japan. These advancements span the fields of control, fabrication, sensing and human-robot interaction, aiming to create robots that are safer, more adaptable, and better integrated into everyday life.
Collision Avoidance in Continuum Robots
A team led by Daniela Rus has developed a closed-form CLF–CBF controller for soft continuum manipulators. This controller guarantees real-time, provably safe 3D collision avoidance without requiring computationally expensive online optimization. The system is designed to scale to numerous obstacles and has been validated through both simulations and physical experiments. MIT CSAIL
Zip-2-Fold: Transformable Robots
Niklas Hagemann and Daniela Rus contributed a novel actuator design called Zip-2-Fold. This 3D-printed actuator can zip and fold into beams with tunable stiffness, enabling the creation of compact robots that can transform into rigid structures. The researchers demonstrated the technology with an adaptive walking robot. MIT CSAIL
AR-Based Teleoperation of Hybrid Robots
Researchers have created an augmented reality (AR)-based framework for teleoperating hybrid rigid-soft robots. Users can control a physical robot through a virtual twin overlaid on the real world. A real-to-sim identification pipeline ensures consistent behavior between the simulation and the physical robot, enhancing safety and intuitiveness during grasping tasks. This perform involves collaboration with researchers at the National University of Singapore and the Singapore-MIT Alliance for Research & Technology M3S program. MIT CSAIL
Adaptive Grasping with Origami Modules
Peiyi Wang and colleagues have designed a multi-finger gripper utilizing passively deformable origami modules. This design allows for adaptive, constant-force grasping with a single actuator and eliminates the need for complex sensing systems. The gripper can grasp multiple stacked objects of varying shapes simultaneously, improving pick-and-place efficiency. MIT CSAIL
Sensorized Soft Robots with Embedded Air Channels
Annan Zhang and a team developed a single-print method for creating helicoid-based soft segments with integrated sensing capabilities. Embedded air channels, pressure sensors, and IMUs provide skin-like perception. The resulting soft arm, reaching a meter in scale with 14 degrees of freedom, is capable of trajectory tracking, grasping, and tactile stiffness detection. The Robot Report
IEEE RoboSoft 2026 Workshop: Minds and Bodies
Alongside the paper presentations, MIT researchers are organizing a workshop titled “Minds and Bodies: Co-Design for the Future of Soft Manipulation” at IEEE RoboSoft 2026. The workshop, taking place on Tuesday, April 7, 2026, in Kanazawa, Japan, features a call for contributions with a deadline of March 21, 2026. A €250 cash prize, sponsored by Compliance Robotics, will be awarded for the best extended abstract and presentation. LinkedIn
The research highlights the growing importance of soft robotics and the collaborative efforts driving innovation in this field. The advancements presented at RoboSoft 2026 demonstrate the potential for soft robots to operate safely and effectively in complex, real-world environments.