Rostok Framework: Automating Underactuated Robot Gripper Design

2026-05-16 · Source: HackerNoon · Field: Technology & Digital — Robotics & Autonomous Systems, Artificial Intelligence & Machine Learning · Depth: Expert, short

Summary

Researchers from ITMO University's Biomechatronics and Energy-Efficient Robotics Lab have developed a novel approach for designing robots, specifically focusing on underactuated tendon-driven grippers. This method integrates morphological computation, which emphasizes achieving robot behavior through body design rather than complex control, with generative design principles. The "rostok" open-source framework was utilized to automatically generate and optimize gripper designs, yielding a set of high-reward solutions. These computationally designed grippers underwent rigorous physical testing to validate their simulated kinematics and confirm their ability to securely grasp objects, demonstrating the potential of automated design in complex robotics.

Key takeaway

For robotics engineers developing specialized grippers, this research suggests a powerful shift from manual design to automated, morphology-driven generation. You should investigate integrating generative design frameworks like "rostok" into your workflow to rapidly prototype and optimize underactuated tendon-driven grippers, potentially reducing development cycles and improving performance through computational exploration of design space.

Key insights

Combining morphological computation and generative design automates the creation of effective robot grippers.

Principles

• Robot behavior can be primarily achieved via body design.
• Automated design can optimize complex robotic components.

Method

The "rostok" pipeline generates tendon-driven gripper designs using morphological computation and generative design, then validates them through physical testing for kinematics and object holding.

In practice

• Explore "rostok" for automated gripper design.
• Prioritize body design for simpler robot control.

Topics

• Rostok Framework
• Underactuated Grippers
• Generative Design
• Morphological Computation
• Tendon-Driven Robots

Best for: AI Scientist, Robotics Engineer, Research Scientist

Related on AIssential

• EAGG: Embodiment-Aligned Grasp Generation via Geometry-Aware Graph Conditioning — Aligning embodiment structure within a shared generator strengthens cross-end-effector grasp generation.
• Vine-inspired robotic gripper gently lifts heavy and fragile objects — Vine-inspired robots can achieve strong, gentle grasps by transforming between open and closed loop configurations.
• CLASP: Language-Driven Robot Skill Selection and Composition using Task-Parameterized Learning — CLASP integrates TP-KMPs with VLMs for data-efficient, language-driven robot skill learning, selection, and composition.
• Leiden scientists 3D-print microrobots that “swim” without a brain — Microrobots can achieve complex navigation through physical design and environmental feedback, not just electronics.
• Periodic Steady-State Control of a Handkerchief-Spinning Task Using a Parallel Anti-Parallelogram Tendon-driven Wrist — A novel tendon-driven wrist and hierarchical control enable precise, periodic manipulation of flexible objects like spinning handkerchiefs.
• RoboLayout: Differentiable 3D Scene Generation for Embodied Agents — RoboLayout generates agent-navigable 3D scenes by integrating explicit reachability constraints and local refinement into differentiable optimization.

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by HackerNoon.