Maximizing Throughput in Multi-Function Robotic Cells

2016-12-12T03:39:31Z (GMT) by Mehdi Foumani
Multi-functionality of robots is almost a new objective of interest, both theoretically and in practice. Recent work has shown the robot is able not only to act as a material handling device but also to inspect the part in transit between machines. Such a kind of robot and the cell in which it is applied are called the <i>Multi-Function Robot</i> (<i>MFR</i>) and the <i>Multi-Function Robotic Cell </i>(<i>MFRC</i>), respectively. Also, the inspection scenario under this condition is named in-line inspection scenario. Considering a MFRC, this thesis contains two main contributions. Firstly, we limit our study to a MFR which only measures the thickness of the part and records results in an independent computer. Accordingly, the processing route of the part is fixed although the MFR performs the inspection process of the part. Under this condition, we find a deterministic model for minimizing the cycle time. Secondly, we consider the user interface computer can be used to modify the processing route of each part based on its inspection result. This means that the number of processing of the part by the production machine is a random variable depending on the inspection result. Consequently, we should develop a stochastic model for minimizing the partial cycle time. For this case, we also focus on two other inspection scenarios in addition to in-line one: post-process and in-process. For the first scenario, the inspection process is performed by an independent inspection machine, while parts are inspected in the production machine using multiple sensors for the second scenario. Since the inspection can be performed by a MFR, we extend results for the in-line scenario. Furthermore, it is shown how a robotic cell with post-process (or in-process) inspection scenario can be converted into a robotic cell with in-line inspection scenario. We propose an analytical method for minimizing cycle time (or expected cycle time) of cells under the aforementioned conditions. Accordingly, the thesis is organized as follows: Chapters 1 and 2 give a general overview of robotic cells, and then Chapters 3-6 present four published papers related to the situation in which the processing route of the part is fixed. The first paper is related to the origin of MFRCs. Following that, second and third papers are related to small- and large-scale MFRCs which only record the inspection results. Finally, the forth paper is related to the operational flexibility in MFRCs. Note that Chapters 3-6 are precedents for Chapters 7-9 where the processing route of each part is modified based on its inspection results. We present two papers in Chapters 7 and 8 to cover robotic cells with post-process and in-process inspection scenarios. Then, in Chapter 9, we show how cells with in-process and post-process inspection scenarios can be converted into a MFRC, which has an in-line inspection scenario. Finally, Chapter 10 presents concluding remarks and some suggestions for MFRCs operating in a dynamic environment.