Foldable robotics is accepted as one of the leading technologies in the soft robotics field. Integrating the sensing components, including hinge angle proprioception, into the robot with a single fabrication method is a part of the field’s ultimate goal. Here we present a cheap single-step method for angle sensing integration into the hinges, with an accurate and reproducible performance. We use silver nanoparticle inkjet printing on the flexible structural layer (PET) of the foldable robot (i.e. Delta robot), using an office-type printer. Silver printed sensors were studied for slight bending applications; however, we report their behavior under a 1 mm minimum radius of curvature, an advanced range both for silver strain sensors and any printed hinge position sensors. Among the three patterns studied, one gave a mean absolute dynamic hysteresis error below 1 degree. Reproducibility of a printed angle sensor behavior is reported for the first time, with three prototypes of each pattern (2degree standard deviation). Printed sensor feedback is tested with proportional control for the first time, via set-point and tracking tasks. On-off control law is also implemented and errors below 1 degree are achieved. Proportional control performances are compared with encoder feedback control and the difference between the realized trajectories are found to be under 1 mm in the task plane.

Accepted and published version of this work can be accessed by  https://doi.org/10.1016/j.sna.2022.113752 . This paper introduces a new method for improved  integrated angle sensing in the foldable robot joints (hinges), using low-cost inkjet printed sensors and a novel physical intelligence based compensation approach. Silver nanoparticle angle sensors are inkjet printed on a flexible PET substrate as a hinge material and implemented as pairs of compression and tension side folding individuals into a rigid experimental setup. Resulting combined signals improved the individual sensor responses due to their compensating physical characteristics, and gave the far best performances in the existing literature in terms of linearity, sensor life-time, static and cyclic drift, hysteresis and dynamic dependency. Proposed approach is promising for eliminating the major limitations on the printed sensor use in flexible hinges and paving the way to reliable, fully soft and easy-to-fabricate all-integrated foldable robots.

In this paper we present the lamination curing as a stand-alone method to activate the silver nanoparticle (Ag NP) inkjet printed angle sensors on a 0.14 mm PET substrate, with a desktop printer. (With the term “lamination curing”, we refer passing the printed sample through a lamination machine, without any actual laminating purpose, only for curing.) We compared the method with the oven curing, which is the widest used method for the intended sensors, and found that lamination cured sensors give lower sheet resistance, lower fabrication uncertainty and more consistent angle sensing behaviors with higher sensing performance. Different curing parameters are inspected and a process under 3 minutes is achieved giving a 0.06 Ohm/square sheet resistance. For such a low sheet resistance, presented method has the lowest thermal curing time among all single layer Ag NP printing studies in the literature. An experimental model is presented for the sheet resistance - aspect ratio relation for both methods. Time dependent resistance shifts of the lamination cured sensors are also inspected and proved to be insignificant. We state lamination curing as an advantageous and reliable alternative to oven curing and other fast curing methods both for sensor and circuitry printing implementations.