Characterization and validation of a novel robotic system for fluid-mediated programmable stochastic self-assembly

Distributed Autonomous Robotic Systems

Khodr

Martinoli

2019

Self Cite

In this paper, we consider a system consisting of multiple floating robotic modules performing self-assembly. Faithfully modeling such a system and its inter-module interactions typically involves capturing the hydrodynamic forces acting on the modules using computationally expensive fluid dynamic modeling tools. This poses restrictions on the usability of the resulting models. Here, we present a new approach towards modeling such systems. First, we show how the hardware and firmware of the robotic modules can be faithfully modeled in a high-fidelity robotic simulator. Second, we develop a physics plugin to recreate the hydrodynamic forces acting on the modules and propose a trajectory-based method for calibrating the plugin model parameters. Our calibration method employs a Particle Swarm Optimization (PSO) algorithm, and consists of minimizing the difference between Mean Squared Displacement (MSD) data extracted from real and simulated trajectories of multiple robotic modules.

Section: Experimental Systemmentioning

confidence: 99%

Design and Calibration of a Lightweight Physics-Based Model for Fluid-Mediated Self-assembly of Robotic Modules

Distributed Autonomous Robotic Systems

Khodr

Martinoli

2019

Self Cite

“…In this paper, we study the SA of our robotic modules in simulation. For the sake of thoroughness, here we give a brief summary of the SA process in our real world experimental system [6]. Our system consists of two main components: 1) the Lily robots, originally presented in [5], which serve as the building blocks of the SA process, and 2) the experimental setup built around them.…”

Section: Fluidic Self-assembly Of Lily Robotsmentioning

confidence: 99%

A Rule Synthesis Algorithm for Programmable Stochastic Self-assembly of Robotic Modules

Distributed Autonomous Robotic Systems

Martinoli

2018

Self Cite

Programmable self-assembly of modular robots offers promising means for structure formation at different scales. Rule-based approaches have been previously employed for distributed control of stochastic self-assembly processes. The assembly rate in the process directly depends on the concurrency level induced by the employed ruleset, i.e. the number of concurrent steps necessary to build one instance of the target structure. Our aim here is to design a formal synthesis algorithm to automatically derive rulesets of high concurrency for a given target structure composed of robotic modules. In the literature, self-assembly of (simulated or real) robotic modules has been realized through manually designed rulesets or manually adjusted rulesets generated by employing graph-grammar formalisms or metaheuristic methods. In this work, we employ an extended graph-grammar formalism, adapted for self-assembly of robotic modules, and propose a novel formal synthesis algorithm capable of generating rulesets for robotic modules by natively considering the morphology of their connectors. The synthesized rulesets induce a high level of concurrency in the self-assembly scheme by exploiting controlled information propagation, using solely local communication. Simulation results of microscopic (non-spatial) and submicroscopic (spatial) models of our robotic platform confirm higher performance of rulesets synthesized by our algorithm compared to related work in the literature.

“…Stochastic SA, demonstrated in [6], [7], [13], [14], is realized through taking advantage of the stochastic ambient energy for module transportation. This in turn allows for further simplification of the modules internal design.…”

Section: Related Workmentioning

confidence: 99%

“…Our system consists of two main components: 1) the floating Lily robotic modules [20], and 2) the experimental setup built around them [7]. The Lilies are endowed with four custom-designed Electro-Permanent Magnets (EPM) to latch and also to communicate locally with their neighbors.…”

Section: Self-assembly In Our Robotic Systemmentioning

confidence: 99%

“…For a given desired target structure, the engineering goal in programmable stochastic self-assembling systems is to derive and program proper ruleset controllers on the modules such that the target structure emerges in a reliable and predictable fashion [5]. The environmental features may also be controlled to assist the assembly of the target structure [6], [7].…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Probabilistic modeling of programmable stochastic self-assembly of robotic modules

2017 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS)

Thandiackal

Mordig

et al. 2017

Self Cite

Abstract-Creating accurate models of stochastically selfassembling systems is a key step in developing control strategies, centralized or distributed, for the self-assembly process. This paper comparatively studies several aspects of developing probabilistic models for programmable self-assembling systems of stochastically interacting modules. In particular, we systematically investigate Markov models as well as hidden Markov models to predict the self-assembly process dynamics. We consider a case study leveraging our fluidic self-assembly robotic system. The ground truth is obtained through a high-fidelity simulation, calibrated using real experimental data. We first consider Markov models and employ the formalism of chemical reaction networks. In order to compute the model parameters, i.e. the reaction rates in the network, three different methods are studied. We then investigate the validity of the underlying well-mixed assumption, and thus the Markov property, for our system through estimation of the diffusion coefficient, through two different approaches. The system is shown to be borderline well-mixed, motivating extension of the initial Markov models to more complex models in order to achieve improved model accuracy. We formulate an automatic method for creating a hidden Markov model starting from a Markov model, based on a previously existing systematic method. Sample trajectories of the models are realized using the Gillespie's method. The resulting hidden Markov model is shown to achieve an improved accuracy over the standard Markov model.