Fits like a Flex-Glove: Automatic Design of Personalized FPCB-Based Tactile Sensing Gloves

TL;DR

The paper tackles the high barrier to creating personalized tactile sensing gloves by introducing an automated pipeline that converts a simple hand photo into FPCB-based resistive sensor designs suitable for board-house manufacturing, achieving per-glove costs under $130 and assembly times under 15 minutes. It employs a hand-landmark–driven design algorithm that places orthogonal electrode grids, generates necessary PCB masks and layers, and outputs production-ready Gerber files, enabling rapid, scalable customization. Experimental characterization shows the sensors exhibit repeatable, pressure-dependent resistance with a linear regime up to ~175 kPa and comparable performance to commercial solutions, while a preliminary user study suggests personalized gloves reduce signal variance and perceived obstruction. The work demonstrates a practical path toward accessible, personalized wearable tactile sensing hardware, while identifying durability and attachment challenges as important directions for future robustness improvements.

Abstract

Resistive tactile sensing gloves have captured the interest of researchers spanning diverse domains, such as robotics, healthcare, and human-computer interaction. However, existing fabrication methods often require labor-intensive assembly or costly equipment, limiting accessibility. Leveraging flexible printed circuit board (FPCB) technology, we present an automated pipeline for generating resistive tactile sensing glove design files solely from a simple hand photo on legal-size paper, which can be readily supplied to commercial board houses for manufacturing. Our method enables cost-effective, accessible production at under \$130 per glove with sensor assembly times under 15 minutes. Sensor performance was characterized under varying pressure loads, and a preliminary user evaluation showcases four unique automatically manufactured designs, evaluated for their reliability and comfort.

Figures (5)

• Figure 1: Automated FPCB design and sensor assembly. (A) Sensing regions are detected and orthogonally arranged electrodes are generated within them. (B) The binary mask of the hand is used to extract its contour, which serves to connect electrodes across neighboring regions. (C) Supporting layers of the PCB are generated and used to create Gerber files which can be sent directly to PCB manufacturers. (D) Sensor assembly using materials totaling under $130 dollars by sandwiching 5 layers together with pressure-sensitive adhesive.
• Figure 2: Sensor Characterizations: (A) Resistance variation during three loading and unloading cycles. (B) Consistent performance is captured by the readout circuit under a load of 12 $N/\text{cm}^2$ for 100 cycles.
• Figure 3: Automatically designed and manufactured FPCB tactile sensors for five individuals. Left: The sensor with median perimeter is used as the basis for a generic glove.
• Figure 4: User study results from generic (Glove A) and personalized (Glove B) gloves. (A-B) Personalized gloves demonstrate reduced variance across repeated full-hand presses. (C) Over 42,00 tactile frames are collected while users perform three tasks—opening a book, using a mouse, and pouring water from a bottle. (D) Users rate personalized gloves as less obstructive and report that tasks require less effort and cause less frustration compared to using generic gloves, on average.
• Figure 5: Mechanical investigation. (A) Failure location on the glove. (B-C) Brittle failure was observed in the electrodeposited (ED) copper design while ductile failure and delamination was observed with the rolled annealed (RA) copper design. (D) Comparison of three designs showing the estimated minimum bending radius before failure. In units of $\mu$m, ED stackup: coverlay-12.5, adhesive-15, polyimide-25 , base ED copper-11.6, and an ENIG overlayer-25 nm, from JLCPCB. RA stackup: coverlay-12.5, adhesive-12.5, polyimide-12.5, base RA copper-12, a nickel overlay-3 and hard gold coating-76 nm, from PCBWay. Encapsulated RA stackup: coverlay-12, adhesive-15, polyimide-25, base RA copper-12, adhesive-15, and coverlay-12, from PCBWay. (E) Manual bending tests show increasing trace width improves RA stackup durability, averaging 330 bends to a tight fist before failure.