We present a strategy for te problem of robotic button mushroom harvesting (Agaricus Bisporus) that involves a Real2Sim2Real pipeline with dynamic scene reconstruction and a Model Predictive Path Integral (MPPI) control & planning architecture for generating optimal uprooting motion primitives based on a physics engine simulation framework. Given the complex, non-linear, anisotropic material properties of the mushrooms in combination with the multiple failure-mode modalities involved, we design a simulation framework around the PyBullet rigid-body-physics engine by utilizing first-order approximations of the equivalent continuum mechanics models. By exploiting the computational efficiency of the aforementioned simulation framework, we directly apply the MPPI control framework to generate offline optimal mushroom uprooting motion primitives, defining a set of cost objectives for an optimal and within-constraint harvesting plan. We show that with this planning strategy, the “root-bending” action emerges autonomously for the case of a single mushroom as an optimal uprooting maneuver, which corresponds well to empirical knowledge obtained by expert pickers. A video demonstration of the proposed architecture can be found in https://youtu.be/k38ePBsBego.

Dynamic stability allows running animals to maintain preferred speed during locomotion over rough terrain. It appears that rapid disturbance rejection is an emergent property of the mechanical system. In running robots, simple motor control seems to be effective in the negotiation of rough terrain when used in concert with a mechanical system that stabilizes passively. In this paper, we show that a quadruped robot could be able to perform self-stable running behavior in significantly broader ranges of forward speed and pitch rate with suitable mechanical design. The results presented here are derived by studying the stability of passive dynamics of a quadruped robot running in the sagittal plane in a dimensionless context and can be summarized as: (a) the self-stabilized behavior of a quadruped robot for a particular gait is related to the magnitude of its dimensionless inertia, (b) the values of hip separation, normalized to rest leg length, and the leg relative stiffness of a quadruped robot affect the stability and should be in inverse proportion to its dimensionless inertia, and (c) the self-stable regime of quadruped running robots is enlarged at relatively high forward speeds.

In legged systems design an important question applies to: "What can be inferred from the performance of a legged robot of a similarly configured system, but scaled to a smaller or larger size?" Our work attempts to answer this question and set the basis for a systematic approach in sizing legged robots. This paper focuses on the influence of permanent magnet DC electric drives on the size of quadruped running robots. The reason is twofold. First, many of the existing legged machines have used such actuators for propulsion. The second, the performance of electric motors scales differently from torque- speed requirements of legged robots. Specifically, we show that there exists a particularly sized quadruped running robot that is superior according to desired performance criteria, and under the existing technologic limitations and economic restraints. Therefore, valuable information on legged systems design and insight for optimizing the size of a quadruped robot emerges.