Instrumentation of biomimetic strain sensors in walking robots
Automation, Mechatronics, Control Systems, Intelligent Technical Systems, Robotics
Systematics and Morphology (Zoology)
Final Report Abstract
Insects have highly robust and reliable locomotion. Through distributed sensors that provide the neuromuscular system with redundant information, they can adapt to rugged conditions and often even the loss of limbs. These sensory structures, that provide critical information about, both, where the limbs are in space, and also, what external factors may be influencing them, allow the control of limb segments to be feedback driven. Robots, on the other hand, are commonly built with simplified sensors that monitor information at one location, to avoid the difficulties of compliance and sensor complexity. This, can be useful in some cases, but in order to have resilient robots that sustain in rugged, unforeseeable terrain, distributed, redundant sensing, like in an insects, may be useful. A type of strain sensor found in insects is called Campaniform sensilla (CS). Through their complex morphology, that is embedded in the cuticle, they can transform strain into neuronal signal. Using a model that is based on a reduced preparation of the these sensilla, that was 3D printed using resins of different compliance, and strain gauge rosettes, we were able to create a mechanical model to test how the mechanical attributes of sensilla can affect strain monitoring in insects. Further, using strain gauge rosettes on dynamically scaled, robotic insect limbs, we could investigate which strains arise at specific time points in the step cycle. The results highlight, that using mechanical models out of resin, nicely captures insect cuticle-like viscoelasticity. Viscoelasticity is thought to lead to adaptation in insect strain sensors which influences their characteristic monitoring of changes in force. Further findings include, that for example, Stick Insect trochanteral CS are more likely to monitor strain at both the anterior and posterior extreme positions, while the homologous CS in the Fruit Fly seem to predominantly monitor the strain only at the anterior extreme position. Besides these more biologically relevant results, our experiments also highlight the relevance of “nontraditional” materials for robotic instrumentation. Resins are proving to have material properties that can be beneficial for the field. Further, implementing even reduced insect-like strain sensor mechanics, such as a more-compliant, higher eccentric component below a strain gauge, can lead to more detailed recordings that demand less computational power. We also investigated insect tarsal structures in robots, the addition of Kevlar in robotic limbs, and the effect of different polymers on strain sensing. In conclusion, we used neurobiological findings and data to create mechanical models of insect strain sensors to test these in physiological loading scenarios. We did this because investigating mechanical displacement in the Fruit Fly is difficult due to, for example, the size and magnitude restraints. Additionally, we used insect-like strain gauges in robotic legs to test distributed strain sensing; implement non-traditional materials for building insect-like robotic legs, and added multicompliant components within these artificial limbs leading to legs with viscoelasticity, directional strain amplification, and local computation of strain.
Publications
-
The BRAIN Initiative Meeting, online, June 11-12, 2023. Abstract title: Trochanteral strain comparison in insects using dynamically scaled robotic legs.
W. P. Zyhowski, G. F. Dinges, C. A. Goldsmith & N. S. Szczecinski
-
Society of Neuroscience Conference, San Diego, Nov 12-16th, 2022. Abstract title: Biomimetic strain sensors in fly-like robot legs.
Gesa F. Dinges, William Zyhowski, Clarissa Goldsmith & Nicholas Szczecinski
-
Subsets of leg proprioceptors influence leg kinematics but not interleg coordination in Drosophila melanogaster walking. Journal of Experimental Biology, 225(20).
Chockley, Alexander S.; Dinges, Gesa F.; Di Cristina, Giulia; Ratican, Sara; Bockemühl, Till & Büschges, Ansgar
-
15th Göttingen Meeting of the German Neuroscience Society, Göttingen, Ger., March 22-24, 2023. Abstract title: Drosophila Strain Sensors: From Morphology and Biomechanics to Function.
Gesa F. Dinges, William Zyhowski & Nicholas Szczecinski
-
Biomimetic and Biohybrid Systems: Living Machines 2023 Conference, Genao, Italy, 2023. Abstract title: Effects of tarsal morphology on load feedback during stepping of robotic stick insect (Carausius morosus) limb.
Clarus Goldsmith, William Zyhowski, Ansgar Büschges, Sasha N. Zill, Gesa F. Dinges & Nicholas Szczecinski
-
Biomimetic and Biohybrid Systems: Living Machines 2023 Conference, Genao, Italy, 2023. Comparison of Proximal Leg Strain in Locomotor Model Organisms Using Robotic legs.
Gesa Dinges, William Zyhowski, Clarus Goldsmith & Nicholas Szczecinski
-
Comparison of Proximal Leg Strain in Locomotor Model Organisms Using Robotic Legs. Lecture Notes in Computer Science, 411-427. Springer Nature Switzerland.
Dinges, Gesa F.; Zyhowski, William P.; Goldsmith, C. A. & Szczecinski, Nicholas S.
-
Effects of Tarsal Morphology on Load Feedback During Stepping of a Robotic Stick Insect (Carausius Morosus) Limb. Lecture Notes in Computer Science, 442-457. Springer Nature Switzerland.
Goldsmith, Clarus A.; Zyhowski, William P.; Büschges, Ansgar; Zill, Sasha N.; Dinges, Gesa F. & Szczecinski, Nicholas S.
-
Lake Conference: Sensation and Action, Thun, Switzerland; May 7-11th, 2023. Abstract title: Biomimetic strain sensors as a tool to investigate load sensing in Drosophila legs.
Gesa F. Dinges, William Zyhowski & Nicholas Szczecinski
-
Panel Discussion: Panellist at “From Animals To Robots: How Transatlantic Neuroscience Advances Engineering” at the German House, 871 UN Plaza, New York City. Nov. 7th, 2023
Ansgar Büschges, Roger D. Quinn, Hillel J. Chiel, Victoria Webster-Wood & Gesa F. Dinges
-
Society of Experimental Biology Centenary Conference 2023, Edinburgh, UK, July 4-7, 2023. Abstract title: Biomimetic Strain Sensors Highlight the Relevance of Arrangement for Load Sensing.
Gesa F. Dinges, William Zyhowski & Nicholas Szczecinski
-
Society of Neuroscience Conference, Washington, D.C., Nov 11-16th, 2023. Abstract title: The Effects of Variability in the Relative Arrangement of Mechanotransducive Sensors on Strain Processing.
Gesa F. Dinges & Nicholas Szczecinski
-
Society of Neuroscience Conference, Washington, D.C., Nov 11-16th, 2023. Abstract title: Using a hexapod robotic model of Drosophila to investigate walking forces and strains.
Clarus Goldsmith, William Zyhowski, Moritz Haustein, Gesa F. Dinges, Ansgar Büschges & Nicholas Szczecinski
-
The BRAIN Initiative Meeting, online, June 11-12, 2023. Abstract title: 3D-Resin Printed Campaniform sensilla as a Tool to Investigate Strain Sensing.
Gesa F. Dinges, William Zyhowski & Nicholas Szczecinski
-
Mechanical modeling of mechanosensitive insect strain sensors as a tool to investigate exoskeletal interfaces. Bioinspiration & Biomimetics, 19(2), 026012.
Dinges, Gesa F.; Zyhowski, William P.; Lucci, Anastasia; Friend, Jordan & Szczecinski, Nicholas S.
