Marvel, Bostelman and Falco (NIST) survey industrial assembly done by two or more robots without fixtures: arms, dexterous hands and mobile platforms such as automated guided vehicles. They sort the literature by the kinds of assemblies that multiple robots make possible, the algorithms that synchronise the robots’ motions, and the metrics used to judge assembly quality and performance.
I only read the abstract; the open author manuscript (PubMed Central PMC5828242) was behind a browser check.
Findings
- From the abstract: the focus is fixtureless assembly, where the robots themselves hold and align parts, which makes coordination and timing between robots part of the task.
- The survey’s other contribution is a catalogue of performance metrics for multi-robot assembly. I have not seen the list.
Limits
- Industrial cells only: structured, mostly static settings with reliable communication.
- I cannot say how much the survey covers safety, fault handling or recovery, since I did not see the body.
Relevance
A useful reference for how a measurement lab like NIST thinks about metrics for multi-robot work, which is relevant to the question of how to score availability and recovery in multi-robot systems and robust robotics.
Links
- DOI · PubMed Central
- Subfield: Multi-robot systems, Robot learning