Modular Reconfigurable Robot Distributed Computing System for Tracking Multiple Objects
- 19 May 2020
- journal article
- research article
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Systems Journal
- Vol. 15 (1) , 802-813
Abstract
A modular reconfigurable robot (MRR) is considered as a system of systems; that is, a robot control system consists of modular systems. For an MRR, onboard tracking of multiple objects is one of the most challenging tasks. This is because the task requires lots of computation resources, but an MRR system is constrained by several physical limitations. To enhance the computing performance of an MRR control system, a novel approach to organizing a distributed computing system is proposed. The key components of the proposed approach are FPGA-based specialized computational cores (SCCs). An SCC for tracking objects is designed to utilize spare resources of the MRR module's control devices. Analysis and experiments on a real testbench are performed to illustrate that the proposed approach can enhance overall computing performance of an MRR control system without adding additional components. As such, weight and dimensional parameters are saved; unnecessary power consumption and heat generation are reduced. One experimental example shows that the proposed SCC enables a Kalman filter to track objects up to 100 times faster than a comparable solution implemented with microcontrollers, while the whole distributed computing system can outperform a similar one built on high-performance CPUs.Keywords
Funding Information
- Russian Foundation for Basic Research (19-38-90301)
- Ministry of Science and Technology, Taiwan (MOST 108-2221-E-007-100)
This publication has 68 references indexed in Scilit:
- Configurable formula calculation tool design based on tagged data structure2015
- A Survey of Research on Cloud Robotics and AutomationIEEE Transactions on Automation Science and Engineering, 2015
- SODA: Software Defined FPGA based Accelerators for Big Data2015
- Computation Sharing in Distributed Robotic Systems: A Case Study on SLAMIEEE Transactions on Automation Science and Engineering, 2014
- Kalman Filter Realization for Orientation and Position Estimation on Dedicated ProcessorActa Mechanica et Automatica, 2014
- A survey on resource allocation in high performance distributed computing systemsParallel Computing, 2013
- A behaviour-based control architecture for heterogeneous modular, multi-configurable, chained micro-robotsRobotics and Autonomous Systems, 2012
- Robotic clusters: Multi-robot systems as computer clustersRobotics and Autonomous Systems, 2012
- MPI-HMMER-Boost: Distributed FPGA AccelerationJournal of Signal Processing Systems, 2007
- Distributed Motion Planning for 3D Modular Robots with Unit-Compressible Modules2004