Search for: All records

The passive compliance of a soft worm-like body can be a key advantage for traversal of complex confined spaces, but in practice, the body’s stiffness and contact friction often require experimental adjustments. Here, for the first time, we develop a dynamic, 3D simulation that enables systematic testing of robot parameters (e.g. stiffness and friction) in different radius of curvature environments, which will help us better understand design trade-offs in creating soft robots that mimic worm-like locomotion. Specifically, we use the open-source physics engine MuJoCo because it is established for both biomechanical and robotic modeling, as well as multi-point contact dynamics, which are present in confined spaces. The model has sensory capabilities analogous to the stretch and tactile proprioception of an earthworm and is amenable to both feedforward and feedback control. After validating our model by comparing to our previous physical robot, we quantify locomotion performance over a range of friction coefficients, structural stiffnesses, and turning radii. We found that speed increased with friction coefficient on flat ground for higher stiffness models, but decreased with friction coefficient for lower stiffness models, both on flat ground and in pipe bends. For turning radii greater than 0.45 m, speed and stiffness also had a positive correlation, however, below the critical turning radius of 0.45 m, increasing stiffness had no appreciable influence on speed. This simulation can potentially be used to optimize designs for particular environments, to better understand the influence of passive vs. active control on individual and coupled segments, and perhaps offer a deeper understanding of how animals and robots can employ soft structures. For example, we can posit from our results that changing stiffness will not increase speed below the critical turning radius, meaning further experiments should focus on other parameters or actively controlled turning to improve speed through tighter turns. 

This paper details the development and validation of a dynamic 3D compliant worm-like robot model controlled by a Synthetic Nervous System (SNS). The model was built and simulated in the physics engine Mujoco which is able to approximate soft bodied dynamics and generate contact, gravitational, frictional, and internal forces. These capabilities allow the model to realistically simulate the movements and dynamic behavior of a physical soft-bodied worm-robot. For validation, the results of this simulation were compared to data gathered from a physical worm robot and found to closely match key behaviors such as deformation propagation along the compliant structure and actuator efficiency losses in the middle segments. The SNS controller was previously developed for a simple 2D kinematic model and has been successfully implemented on this 3D model with little alteration. It uses coupled oscillators to generate coordinated actuator control signals and induce peristaltic locomotion. This model will be useful for analyzing dynamic effects during peristaltic locomotion like contact forces and slip as well as developing and improving control algorithms that avoid unwanted slip. 

One developing approach for robotic control is the use of networks of dynamic neurons connected with conductance-based synapses, also known as Synthetic Nervous Systems (SNS). These networks are often developed using cyclic topologies and heterogeneous mixtures of spiking and non-spiking neurons, which is a difficult proposition for existing neural simulation software. Most solutions apply to either one of two extremes, the detailed multi-compartment neural models in small networks, and the large-scale networks of greatly simplified neural models. In this work, we present our open-source Python package SNS-Toolbox, which is capable of simulating hundreds to thousands of spiking and non-spiking neurons in real-time or faster on consumer-grade computer hardware. We describe the neural and synaptic models supported by SNS-Toolbox, and provide performance on multiple software and hardware backends, including GPUs and embedded computing platforms. We also showcase two examples using the software, one for controlling a simulated limb with muscles in the physics simulator Mujoco, and another for a mobile robot using ROS. We hope that the availability of this software will reduce the barrier to entry when designing SNS networks, and will increase the prevalence of SNS networks in the field of robotic control. 

This work explored synaptic strengths in a computational neuroscience model of a controller for the hip joint of a rat which consists of Ia interneurons, Renshaw cells, and the associated motor neurons. This circuit has been referred to as the Canonical Motor Microcircuit (CMM). It is thought that the CMM acts to modulate motor neuron activity at the output stage. We first created a biomechanical model of a rat hindlimb consisting of a pelvis, femur, shin, foot, and flexor-extensor muscle pairs modeled with a Hill muscle model. We then modeled the CMM using non-spiking leaky-integrator neural models connected with conductance-based synapses. To tune the parameters in the network, we implemented an automated approach for parameter search using the Markov chain Monte Carlo (MCMC) method to solve a parameter estimation problem in a Bayesian inference framework. As opposed to traditional optimization techniques, the MCMC method identifies probability densities over the multidimensional space of parameters. This allows us to see a range of likely parameters that produce model outcomes consistent with animal data, determine if the distribution of likely parameters is uni- or multi-modal, as well as evaluate the significance and sensitivity of each parameter. This approach will allow for further analysis of the circuit, specifically, the function and significance of Ia feedback and Renshaw cells.