# Method for Robot Manipulator Joint Wear Reduction by Finding the Optimal Robot Placement in a Robotic Cell

^{1}

^{2}

^{3}

^{*}

## Abstract

**:**

## 1. Introduction

#### The Concept of Robot Wear

## 2. Materials and Methods

#### 2.1. Optimization Criterion

- to minimize the overall wear of all joints, and
- to balance the wear of all joints.

#### 2.2. Optimization Process

#### 2.3. Experiment Setup

## 3. Results

## 4. Verification on a Real Robot

## 5. Discussion

## 6. Conclusions

## Author Contributions

## Funding

## Institutional Review Board Statement

## Informed Consent Statement

## Data Availability Statement

## Conflicts of Interest

## Abbreviations

API | Application programming interface |

CSO | Chicken swarm optimization |

GCS | Global coordinate system |

IK | Inverse kinematics |

PSO | Particle swarm optimization |

RTDE | Real-time data exchange |

TCP/IP | Transmission control protocol/Internet protocol |

UR3 | Universal Robots 3 |

## References

- Diaz, J.R.; Mukherjee, P.; Verl, A. Automatic Close-optimal Workpiece Positioning for Robotic Manufacturing. Procedia CIRP
**2018**, 72, 277–284. [Google Scholar] [CrossRef] - Tangpattanakul, P.; Artrit, P. Minimum-time trajectory of robot manipulator using Harmony Search algorithm. In Proceedings of the 2009 6th International Conference on Electrical Engineering/Electronics, Computer, Telecommunications and Information Technology, Chonburi, Thailand, 6–9 May 2009; Volume 1, pp. 354–357. [Google Scholar] [CrossRef]
- Valente, A.; Baraldo, S.; Carpanzano, E. Smooth trajectory generation for industrial robots performing high precision assembly processes. CIRP Ann.
**2017**, 66, 17–20. [Google Scholar] [CrossRef] - Llopis-Albert, C.; Rubio, F.; Valero, F. Modelling an Industrial Robot and Its Impact on Productivity. Mathematics
**2021**, 9, 769. [Google Scholar] [CrossRef] - Piazzi, A.; Visioli, A. Global minimum-jerk trajectory planning of robot manipulators. IEEE Trans. Ind. Electron.
**2000**, 47, 140–149. [Google Scholar] [CrossRef] [Green Version] - Gasparetto, A.; Zanotto, V. A technique for time-jerk optimal planning of robot trajectories. Robot. Comput. Integr. Manuf.
**2008**, 24, 415–426. [Google Scholar] [CrossRef] - Hu, S.; Kang, H.; Tang, H.; Cui, Z.; Liu, Z.; Ouyang, P. Trajectory Optimization Algorithm for a 4-DOF Redundant Parallel Robot Based on 12-Phase Sine Jerk Motion Profile. Actuators
**2021**, 10, 80. [Google Scholar] [CrossRef] - Mu, Y.; Zhang, L.; Chen, X.; Gao, X. Optimal Trajectory Planning for Robotic Manipulators Using Chicken Swarm Optimization. In Proceedings of the 2016 8th International Conference on Intelligent Human-Machine Systems and Cybernetics (IHMSC), Hangzhou, China, 27–28 August 2016; Volume 2, pp. 369–373. [Google Scholar] [CrossRef]
- Lim, Z.Y.; Ponnambalam, S.G.; Izui, K. Nature inspired algorithms to optimize robot workcell layouts. Appl. Soft Comput.
**2016**, 49, 570–589. [Google Scholar] [CrossRef] - Bukata, L.; Šůcha, P.; Hanzálek, Z.; Burget, P. Energy Optimization of Robotic Cells. IEEE Trans. Ind. Inform.
**2017**, 13, 92–102. [Google Scholar] [CrossRef] - Gadaleta, M.; Berselli, G.; Pellicciari, M. Energy-optimal layout design of robotic work cells: Potential assessment on an industrial case study. Robot. Comput. Integr. Manuf.
**2017**, 47, 102–111. [Google Scholar] [CrossRef] - Pellicciari, M.; Berselli, G.; Leali, F.; Vergnano, A. A method for reducing the energy consumption of pick-and-place industrial robots. Mechatronics
**2013**, 23, 326–334. [Google Scholar] [CrossRef] - Meike, D.; Pellicciari, M.; Berselli, G. Energy Efficient Use of Multirobot Production Lines in the Automotive Industry: Detailed System Modeling and Optimization. IEEE Trans. Autom. Sci. Eng.
**2014**, 11, 798–809. [Google Scholar] [CrossRef] - Paes, K.; Dewulf, W.; Elst, K.V.; Kellens, K.; Slaets, P. Energy Efficient Trajectories for an Industrial ABB Robot. Procedia CIRP
**2014**, 15, 105–110. [Google Scholar] [CrossRef] [Green Version] - Spensieri, D.; Carlson, J.S.; Bohlin, R.; Kressin, J.; Shi, J. Optimal Robot Placement for Tasks Execution. Procedia CIRP
**2016**, 44, 395–400. [Google Scholar] [CrossRef] - Biesinger, F.; Meike, D.; Krass, B.; Weyrich, M. A Case Study for a Digital Twin of Body-in-White Production Systems General Concept for Automated Updating of Planning Projects in the Digital Factory. In Proceedings of the 2018 IEEE 23rd International Conference on Emerging Technologies and Factory Automation (ETFA), Turin, Italy, 4–7 September 2018; Volume 1, pp. 19–26. [Google Scholar] [CrossRef]
- Miková, Ľ.; Kelemen, M.; Virgala, I.; Michna, M. Simulation Model of Manipulator for Model Based Design. Appl. Mech. Mater.
**2014**, 611, 175–182. [Google Scholar] [CrossRef] - Georgoulas, G.; Loutas, T.; Stylios, C.D.; Kostopoulos, V. Bearing fault detection based on hybrid ensemble detector and empirical mode decomposition. Mech. Syst. Signal Process.
**2013**, 41, 510–525. [Google Scholar] [CrossRef] - Qian, H.M.; Li, Y.F.; Huang, H.Z. Time-variant reliability analysis for industrial robot RV reducer under multiple failure modes using Kriging model. Reliab. Eng. Syst. Saf.
**2020**, 199, 106936. [Google Scholar] [CrossRef] - Guangjian, W.; Lin, C.; Li, Y.; Shuaidong, Z. Research on the dynamic transmission error of a spur gear pair with eccentricities by finite element method. Mech. Mach. Theory
**2017**, 109, 1–13. [Google Scholar] [CrossRef] - Cubillo, A.; Perinpanayagam, S.; Esperon-Miguez, M. A review of physics-based models in prognostics: Application to gears and bearings of rotating machinery. Adv. Mech. Eng.
**2016**, 8, 1687814016664660. [Google Scholar] [CrossRef] [Green Version] - Zhang, X.; Jiang, G.; Zhang, H.; Yun, X.; Mei, X. Time-dependent reliability analysis of harmonic drive based on transient FEA and accelerated life test. Eng. Comput.
**2020**. [Google Scholar] [CrossRef] - Bian, L.; Gebraeel, N.; Kharoufeh, J.P. Degradation modeling for real-time estimation of residual lifetimes in dynamic environments. IIE Trans.
**2015**, 47, 471–486. [Google Scholar] [CrossRef] - Koike, H.; Itakura, K.; Okazaki, S.; Takamiya, M.; Kanemasu, K.; Kida, K. Measurement of Backlash and Fatigue Wear of PEEK Bush in Robot Joint under Middle Load. In Applied Mechanics and Materials; Trans Tech Publications Ltd.: Stafa-Zurich, Switzerland, 2013; Volume 418, pp. 38–43. [Google Scholar] [CrossRef]
- Bittencourt, A.C.; Axelsson, P.; Jung, Y.; Brogårdh, T. Modeling and Identification of Wear in a Robot Joint under Temperature Uncertainties. IFAC Proc. Vol.
**2011**, 44, 10293–10299. [Google Scholar] [CrossRef] [Green Version] - Lugt, P.M. A Review on Grease Lubrication in Rolling Bearings. Tribol. Trans.
**2009**, 52, 470–480. [Google Scholar] [CrossRef] - Aivaliotis, P.; Arkouli, Z.; Georgoulias, K.; Makris, S. Degradation curves integration in physics-based models: Towards the predictive maintenance of industrial robots. Robot. Comput. Integr. Manuf.
**2021**, 71, 102177. [Google Scholar] [CrossRef] - Peternel, L.; Tsagarakis, N.; Ajoudani, A. A Method for Robot Motor Fatigue Management in Physical Interaction and Human-Robot Collaboration Tasks. In Proceedings of the 2018 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), Madrid, Spain, 1–5 October 2018; pp. 2850–2856. [Google Scholar] [CrossRef]
- Rodrigues, L.R. Remaining Useful Life Prediction for Multiple-Component Systems Based on a System-Level Performance Indicator. IEEE/ASME Trans. Mechatron.
**2018**, 23, 141–150. [Google Scholar] [CrossRef] - Jacobs, S.; Rios-Gutierrez, F. Self organizing maps for monitoring parameter deterioration of DC and AC motors. In Proceedings of the 2013 Proceedings of IEEE Southeastcon, Jacksonville, FL, USA, 4–7 April 2013; pp. 1–6. [CrossRef] [Green Version]
- Poli, R.; Kennedy, J.; Blackwell, T. Particle swarm optimization. Swarm Intell.
**2007**, 1, 33–57. [Google Scholar] [CrossRef] - Kovincic, N.; Mueller, A.; Gattringer, H.; Weyrer, M.; Schlotzhauer, A.; Brandstötter, M. Dynamic parameter identification of the Universal Robots UR5. In Proceedings of the Austrian Robotics Workshop 2019, Steyr, Austria, 9–10 May 2019. [Google Scholar] [CrossRef]

**Figure 1.**Example of the time progression of joint torque $\tau $ and angular velocity $\omega $; the dotted area represents the meaning of the value W calculated as the integral of absolute value $\left(\right)$.

**Figure 2.**The UR3 robot. (

**a**) The simulation model in the CoppeliaSim environment with a path representing the end-point trajectory and a coordinate system in the center of the robot base. (

**b**) The real robot used in the experiments.

**Figure 3.**Visualization of the five testing robot end-point paths; the UR3 robot is shown as a scale reference.

**Figure 4.**Example of a grid of valid robot locations for a given trajectory (A1). Each blue point represents a possible position of the robot base center point; the robot is displayed in four distinct sample locations. GCS represents the global coordinate system.

**Figure 5.**Statistical distribution of the ${f}_{f}$ values in all possible robot locations for the ten testing trajectories (standard box-plot diagram—each box is bounded by the first and third quartile, the horizontal line represents the median, the × represents the arithmetic mean, and circles represent outliers).

**Figure 6.**All valid robot locations in the grid; the color-coding indicates the ${f}_{f}$ values using the gradient red–yellow–green, where green is the best location (B) and red is the worst (W). Locations numbered 1, 2, and 3 are the three other locations used in the experiment with a real robot.

**Figure 7.**Detailed comparison of the components forming the fitness function value in the best (B, green) and worst (W, red) robot locations for two selected trajectories; displayed are the relative wear factors of each joint (${w}_{j}$) and the total fitness function value (${f}_{f}$).

**Figure 8.**Comparison of the angular velocity ${\omega}_{3}$, torque ${\tau}_{3}$, and immediate power ${\omega}_{3}{\tau}_{3}$ of the third joint during the whole trajectories A1 and B1 for the best (${}^{B}$) and worst (${}^{W}$) robot locations.

**Figure 9.**Comparison of the fitness function values acquired from simulation and experiments on the real robot; the black lines represent the ratio between the values.

**Figure 10.**Comparison of the real and simulated values of angular velocity ${\omega}_{3}$ and torque ${\tau}_{3}$ of the third joint during the whole trajectory B1. (

**a**) Robot placed in the best location. (

**b**) Robot placed in the worst location.

**Table 1.**The joint parameters for the UR3 robot; ${\tau}_{{m}_{j}}$ is the maximal permissible torque, and ${\omega}_{{m}_{j}}$ is the maximal permissible angular velocity of the j-th joint.

j | ${\mathit{\tau}}_{{\mathit{m}}_{\mathit{j}}}$ [Nm] | ${\mathit{\omega}}_{{\mathit{m}}_{\mathit{j}}}$ [s^{−1}] |
---|---|---|

1 | 56 | $\pi $ |

2 | 56 | $\pi $ |

3 | 28 | $\pi $ |

4 | 12 | $2\pi $ |

5 | 12 | $2\pi $ |

6 | 12 | $2\pi $ |

**Table 2.**Results for the 10 testing trajectories showing the number of valid robot locations n, the fitness function value in the best location ${{f}_{f}}^{B}$, in the worst location ${{f}_{f}}^{W}$, the average value ${{f}_{f}}^{A}$; and improvement of the best location against the worst $im{p}_{W}^{B}$ (9) and the average $im{p}_{A}^{B}$ location (10).

Traj. | n | ${{\mathit{f}}_{\mathit{f}}}^{\mathit{B}}$ | ${{\mathit{f}}_{\mathit{f}}}^{\mathit{W}}$ | ${{\mathit{f}}_{\mathit{f}}}^{\mathit{A}}$ | ${\mathit{i}\mathit{m}\mathit{p}}_{\mathit{W}}^{\mathit{B}}$ | ${\mathit{i}\mathit{m}\mathit{p}}_{\mathit{A}}^{\mathit{B}}$ |
---|---|---|---|---|---|---|

A1 | 800 | 0.0730 | 0.2053 | 0.1239 | 64.4% | 41.1% |

A2 | 564 | 0.0773 | 0.1577 | 0.1046 | 51.0% | 26.1% |

B1 | 2265 | 0.0217 | 0.1090 | 0.0461 | 80.1% | 52.9% |

B2 | 2236 | 0.0210 | 0.1067 | 0.0448 | 80.3% | 53.1% |

C1 | 2365 | 0.0769 | 0.1998 | 0.1239 | 61.5% | 37.9% |

C2 | 2214 | 0.0757 | 0.1828 | 0.1170 | 58.6% | 35.2% |

D1 | 866 | 0.0907 | 0.1994 | 0.1163 | 54.5% | 22.0% |

D2 | 828 | 0.0841 | 0.1829 | 0.1097 | 54.0% | 23.3% |

E1 | 2224 | 0.0279 | 0.1028 | 0.0532 | 72.8% | 47.4% |

E2 | 2151 | 0.0247 | 0.0984 | 0.0502 | 74.9% | 50.7% |

**Table 3.**Fitness function values in the five robot locations for individual trajectories—results from the simulation. Color gradient red–yellow–green is used for better visual representation of the value (green is the best, red is the worst).

B | 1 | 2 | 3 | W | |
---|---|---|---|---|---|

A1 | 0.0730 | 0.1256 | 0.1281 | 0.1241 | 0.2054 |

A2 | 0.0774 | 0.1025 | 0.1156 | 0.1434 | 0.1578 |

B1 | 0.0217 | 0.0357 | 0.0389 | 0.0484 | 0.1090 |

B2 | 0.021 | 0.0289 | 0.0310 | 0.0588 | 0.1068 |

C1 | 0.0770 | 0.1074 | 0.1195 | 0.1421 | 0.1998 |

C2 | 0.0758 | 0.0939 | 0.1040 | 0.1160 | 0.1828 |

D1 | 0.0907 | 0.1045 | 0.1077 | 0.1187 | 0.1995 |

D2 | 0.0841 | 0.0967 | 0.1157 | 0.1203 | 0.1830 |

E1 | 0.028 | 0.0358 | 0.0388 | 0.0642 | 0.1028 |

E2 | 0.0247 | 0.0371 | 0.0585 | 0.0606 | 0.0985 |

**Table 4.**Fitness function values in the five robot locations for individual trajectories–results from the real robot. Color gradient red–yellow–green is used for better visual representation of the value (green is the best, red is the worst).

B | 1 | 2 | 3 | W | |
---|---|---|---|---|---|

A1 | 0.0506 | 0.0799 | 0.0815 | 0.0848 | 0.1343 |

A2 | 0.054 | 0.0752 | 0.0799 | 0.0900 | 0.1155 |

B1 | 0.019 | 0.0312 | 0.0317 | 0.0278 | 0.0681 |

B2 | 0.0191 | 0.0263 | 0.0280 | 0.0348 | 0.0679 |

C1 | 0.0635 | 0.0760 | 0.0828 | 0.0874 | 0.1176 |

C2 | 0.0621 | 0.0625 | 0.0810 | 0.0808 | 0.1124 |

D1 | 0.078 | 0.0877 | 0.0940 | 0.0783 | 0.1015 |

D2 | 0.0777 | 0.0869 | 0.1038 | 0.0857 | 0.0958 |

E1 | 0.0239 | 0.0310 | 0.0348 | 0.0457 | 0.0599 |

E2 | 0.025 | 0.0355 | 0.0481 | 0.0446 | 0.0621 |

**Table 5.**Ratios between the real and simulated values of fitness function in the five robot locations for individual trajectories. Color gradient red–white is used for better visual representation of the value (white is the ideal ratio of 1.0, red is the ratio 0.5).

B | 1 | 2 | 3 | W | |
---|---|---|---|---|---|

A1 | 0.6932 | 0.6359 | 0.6362 | 0.6833 | 0.6541 |

A2 | 0.6985 | 0.7334 | 0.6912 | 0.6278 | 0.732 |

B1 | 0.8743 | 0.8722 | 0.8168 | 0.5735 | 0.6244 |

B2 | 0.9056 | 0.9096 | 0.9025 | 0.5915 | 0.6359 |

C1 | 0.8246 | 0.7075 | 0.6933 | 0.6151 | 0.5887 |

C2 | 0.8199 | 0.6657 | 0.7792 | 0.6965 | 0.6149 |

D1 | 0.8594 | 0.8393 | 0.8728 | 0.6591 | 0.5087 |

D2 | 0.9243 | 0.8984 | 0.8968 | 0.7126 | 0.5238 |

E1 | 0.8541 | 0.8658 | 0.8960 | 0.7118 | 0.5825 |

E2 | 1.0100 | 0.9590 | 0.8224 | 0.7360 | 0.6303 |

Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations. |

© 2021 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https://creativecommons.org/licenses/by/4.0/).

## Share and Cite

**MDPI and ACS Style**

Kot, T.; Bobovský, Z.; Vysocký, A.; Krys, V.; Šafařík, J.; Ružarovský, R.
Method for Robot Manipulator Joint Wear Reduction by Finding the Optimal Robot Placement in a Robotic Cell. *Appl. Sci.* **2021**, *11*, 5398.
https://doi.org/10.3390/app11125398

**AMA Style**

Kot T, Bobovský Z, Vysocký A, Krys V, Šafařík J, Ružarovský R.
Method for Robot Manipulator Joint Wear Reduction by Finding the Optimal Robot Placement in a Robotic Cell. *Applied Sciences*. 2021; 11(12):5398.
https://doi.org/10.3390/app11125398

**Chicago/Turabian Style**

Kot, Tomáš, Zdenko Bobovský, Aleš Vysocký, Václav Krys, Jakub Šafařík, and Roman Ružarovský.
2021. "Method for Robot Manipulator Joint Wear Reduction by Finding the Optimal Robot Placement in a Robotic Cell" *Applied Sciences* 11, no. 12: 5398.
https://doi.org/10.3390/app11125398