Everyday Finger: A Robotic Finger that Meets the Needs of Everyday Interactive Manipulation

TL;DR

This work addresses the challenge of dexterous manipulation in everyday environments by proposing a compact, compliant two-finger hand driven by series-elastic actuators. It defines five performance metrics—compactness, compliance, force, speed, and bandwidth—and uses a data-driven, optimization-based approach to map human daily-tool interactions to joint torque and bandwidth requirements, guiding the SEA-based design. The Everyday Finger employs a three-stage transmission with an embedded elastic element and a 4-bar linkage, using off-the-shelf motors to achieve notable torque (MCP ≈ 1.87 Nm, PIP ≈ 0.68 Nm), high speeds (up to ~21.6 rad/s for PIP), and good backdrivability, validated through both quantitative characterization and qualitative tasks (dish handling, napkin pickup, strawberry grasping). While the knuckle footprint exceeds the target in some dimensions, the results demonstrate a viable path toward compact, affordable, everyday manipulation hardware and point to future work on 3–4 finger configurations and tactile sensing to enable full-hand dexterity.

Abstract

We provide the mechanical and dynamical requirements for a robotic finger capable of performing thirty diverse everyday tasks. To match these requirements, we present a finger design based on series-elastic actuation that we call the everyday finger. Our focus is to make the fingers as compact as possible while achieving the desired performance. We evaluated everyday fingers by constructing a two-finger robotic hand that was tested on various performance parameters and tasks like picking and placing dishes in a rack, picking thin and flat objects like paper and delicate objects such as strawberries. Videos are available at the project website: https://sites.google.com/view/everydayfinger.

Figures (14)

• Figure 1: The proposed design of the everyday two-finger robotic hand that is appropriate for daily tasks.
• Figure 2: Left: Finger geometry and degrees of freedom used in optimization. Left-Middle: diagram of three contact points (orange) on a cylindrical tool handle and their corresponding forces as used in optimization. Right: Our hand's capabilities compared to daily task requirements. Tasks are split between small, medium and large handle radius, as well as whether 4th contact paint, a palm, was used.
• Figure 3: Medium Pinch grasp, such as is used for the screwdriver tasks, for inferring torque trajectories at the joints based on FT sensor readings (shown in red in the diagram).
• Figure 4: Optimized peeling cucumber joint torque trajectory for index finger MCP-X joint and the response of a first order system with 5Hz bandwidth, the requirement for this task based on our analysis. Note that the high frequency noise from the raw data is attenuated, but the main signal is followed closely.
• Figure 5: Classic model for a series elastic actuator.