Phd thesis in robotics and mechanical engineering

Updated: 30 days ago
Location: Tremblay en France, LE DE FRANCE
Deadline: 09 Sep 2019

The doctoral thesis will take place at the Laboratoire des Sciences du Numérique de Nantes (LS2N - UMR 6004), Nantes Central Campus.

The doctoral student will have at his or her disposal a cell comprising a cable-driven parallel robot and several measuring instruments for the experimental part of his or her research work.

%%%
Prerequisites:

The candidate must hold an engineering degree and/or a master's degree in advanced mechanical engineering or robotics. The position requires a good knowledge of:

1) Mechanical design
2) Modeling of polyarticulated and flexible systems
3) Robotic system control

as well as

- good oral and written communication skills (writing articles, presenting at meetings or congresses, interactions)
- a very good level in English writing and speaking (TOEIC level > 900)
- an ability to work in a team

Cable-Driven Parallel Robots (CDPRs) form a particular class of parallel robots whose moving platform is connected to a fixed base frame by cables. The cables may be coiled/uncoiled by motorized winches allowing a control system to adjust the cable lengths between the winch exit points and the cable attachment points on the platform. Appropriate length adjustment of N cables allows one to control N degrees-of-freedom of the moving platform. CDPRs have several advantages such as mechanical simplicity and a potentially very large workspace. They have been used in several applications, e.g. heavy payload handling and airplane painting [ABD92], cargo handling [HC04], warehouse applications [HK09], large-scale assembly and handling [PMV10], manipulation in hazardous environment [R+04], fast pick-and-place operations [KKW00], broadcasting of sporting events, haptic devices [FCG14], giant telescopes [YTWH10], environment monitoring [JBK07] and search and rescue platforms [MD10]. When CDPR are working together with human operators, safety issues, usability and acceptability have to be addressed as part of the design process.

The performance improvement of CDPRs is still a challenge because the cable unilaterality (cables can pull but cannot push) makes their analysis much more complex than parallel robots with rigid legs. Indeed, it is necessary to introduce well known cable models in the analysis, that are usually complex, non-linear, possibly non-algebraic and compute-intensive. To simplify this modeling, it will be interesting to introduce additional sensing beside the measurement of the cable lengths, so that sensor fusion can be used to improve the CDPR performance. The resulting system will allow self-calibration, so that wear and characteristic changes are taken into account in the control. To address those challenges we will develop an appropriate software framework of numerical solving methods, whose validity will be verified on realistic experiments. The integration of theoretical modeling, numerical solving and experiments will result in strong technological and scientific advances on the topic. Besides accurate solving and self-calibration are tremendous to provide information to the operators about robot operations through specific Human Machine Interfaces.

As previously mentioned, CDPRs may perform various kinds of applications but their performances are strongly dependent upon their geometry. This strong dependency makes the analysis complex and could be seen as a drawback. However, it allows one to produce CDPRs with a large variety of performances that are based on same hardware but various geometries, paving the way to choose the CDPRs that are appropriate for the task requirements and the environment. We aim therefore at developing agile CDPRs, called Reconfigurable Cable-Driven Parallel Robots (RCDPRs), that can be adapted to the task. From a mechanical point of view, changing the CDPR geometry is relatively easy either by moving the winches, by using pulleys to modify the location of the winch exit points or by changing the location of the cable attachment points on the platform. Preliminary studies on RCDPRs were performed in the context of the NIST RoboCrane project [BJPK00]. The real challenge for RCDPRs is to develop a strategy for determining the best CDPR geometries for a given task to be fulfilled, considering the unavoidable uncertainties in the system. [NGCP14] proposed reconfiguration strategies for large-dimension suspended CDPRs mounted on overhead bridge cranes. Discrete reconfigurations where the locations of the cable exit points are selected from a finite set have been recently studied in [GCGG16] while the general case has been addressed in [Bla15]. We aim at developing a design framework that takes as inputs the task requirements and a description of the CDPR environment. It will provide as output a set of possible CDPR geometries together with main performance indicators whose calculation will also take into account the manufacturing uncertainties.

The work that will be carried out in the framework of this PhD thesis aims to design, model and control CDPRs for agile operations in manufacturing facilities. The CDPRs developed in the framework of this PhD thesis should be able to work in a cluttered environment and assist operators in carrying and manipulating large and heavy parts. Therefore, the CDPRs under study will have two working modes. The robots will either be autonomous (with possibly some human operators in its environment) and will realize some tasks that are set up offline or the user will safely co-manipulate large and/or heavy objects with the cobot in large and cluttered environments. The robots developed in the framework of the project will provide the user with a good agility in large workspace while ensuring the safety of the human operator. Here, the motions of the objects are supposed to be slow and adapted to the identified risks.

The PhD thesis work will be broken down into the following tasks:

1) Design strategy of reconfigurable CDPRs: The robots studied in the framework of project will be reconfigurable because they should be easy to deploy and install and should be adapted to the task to be realized and the environment. Reconfigurability is mostly based on changes in the CDPRs geometry and has already been considered either uniquely for the hardware part [I+12] or on the algorithmic side [B+17],[Bla15],[Gag16]. However, it has not considered complex cable models as presented in the next section and while neglecting some important points such as stiffness ,[KB10],[MOr13], energy consumption, stability [CM13] and the singularities (poses in which there is a loss of control of the CDPR and that have never been studied for CDPRs with sagging cables). Evolutionary algorithms have been used in CDPR design [H+15] but without considering the effect of uncertainties. Dynamic programming and interval analysis seem to be more appropriate to determine the best design solutions while providing guarantee about the manufactured CDPR performance [A+10],[FM05]. Those methods will be also used to determine the best kinematic redundancy scheme of agile, easy to install CDPRs.

2) Kineto-static models of CDPRs: Determining the relationships between cable lengths and platform pose is crucial for controlling CDPRs. Issues such as cable sagging and the consideration of pulleys have been addressed in the past [GNB14], [Pot12], [R+09], [R+10a], [R+10b], [K+06], [KKP15], [M+14] both in theory and experimentally but under strong simplifying assumptions [N+13], [SWI17]. The consideration of a more realistic cable model has been shown to be quite difficult and has provided results with significant differences compared to simplified cable models [S+09], [Mer16a], [Mer16b]. Another important issue is related to the influence of the discrete time control of CDPRs which induces unexpected changes in cable tensions compared to the one obtained through continuous time simulation [Mer17]. Such an analysis, which is numerically demanding, has to be extended to more realistic cable models, especially for large robots. A possibility to tackle in real-time the kineto-static problems of CDPRs with realistic cable models is to instrument the system with proprioceptive sensors and to fuse them with the cable lengths in order to get a better estimation of the CDPR state. Preliminary studies have shown that indeed measuring the cable orientation (for example through a vision system [D+11] although this method may not be appropriate for large outdoor CDPRs) may provide interesting data. However uncertainties in the measurements have not been taken into account and other sensing possibilities (accelerometers, lidar,..) have not been considered. Consequently an analysis of the influence of uncertainties will be performed and will be confirmed by experimental results in the scope of this PhD thesis.

3) Identification of CDPR physical parameters: Solving the kineto-static models and designing an efficient model-based control requires an accurate estimation of CDPR parameters such as geometry, elasticity of the cables. Calibration is therefore required and self-calibration, that requires only the use of the proprioceptive sensor of the CDPR, will be favored. This subject has been addressed in [AdSDG12], [C+04], [MP12] but under the assumption of non-deformable cables and without considering the additional sensors described in the previous section [JS03]. Therefore, original methods will be developed in this PhD thesis to identify those parameters either during the robot operation whenever possible or during the maintenance phases. Sensitivity analysis is a prerequisite for self-calibration for analyzing the parameter identifiability. This analysis will provide a catalogue of possible sensors configuration (together with the corresponding identifiable parameters, and the relationship between the parameters and sensors errors) and the computing time of the different self-calibration procedures. Based on these elements we may either store calibration data during the normal operation of the CDPR for a posteriori calibration or use on-line calibration. This analysis will also be used in the design phase for providing information on the necessary sensory hardware according to the task at hand.

4) CDPR control strategies: CDPRs have the property that their state equations change according to the robot configuration (e.g if a cable becomes slack, its influence on the robot behavior will be minimal). Hence adaptive control schemes are required and will be synthesized in the CRAFT project. Furthermore, we will develop control strategies that allow the robot to switch from an “autonomous” mode to “co-manipulation” one in which the user can interact manually with the robot without feeling the payload, while the safety of the user is guaranteed. The control strategies should ensure safety of in both modes while avoiding the collisions between the cables, the platform and the user [Bla15]. Furthermore, these control strategies should be accepted by the human operators as mentioned in the next section.

References :
[A+10] S. Abdelaziz et al. Combining structural and kinematic analysis using interval analysis for a wire-driven manipulator. In ARK, pages 147–156, Piran, June 28- July 1, 2010.
[ABD92] Albus, J., Bostelman, R., Dagalakis, N.: The NIST spider, a robot crane. J. of Research of the Nat. Inst. of Standards and Technology 97(3), 373–385 (1992)
[AdSDG12] J. Alexandre dit Sandretto et al. Calibration of a fully-constrained parallel cable-driven robot. RoManSy,, Paris, June, 12-15, 2012.
[A14] Acoulon, S., “Les robots collaboratifs : guide d'intégration de la santé et de la sécurité.”, Coll. Les ouvrages du CETIM, 2014
[B+17] L. Barbazza et al. Trajectory planning of a suspended cable driven parallel robot with reconfigurable end effector. Robotics and Computer-Integrated Manufacturing, 48(6):1–11, 2017.
[Bla15] L. Blanchet. Contribution à la modélisation de robots à câbles pour leur commande et leur conception. PhD thesis, Université de Nice- Sophia Antipolis, Nice, May, 13, 2015.
[BJPK00] Bostelman R. et al: Cable-based reconfigurable machines for large scale manufacturing. Japan-USA Symp. on Flexible Automation,2000
[CCCG15] A. L. Cruz Ruiz, S. Caro, P. Cardou, and F. Guay, “ARACHNIS : Analysis of Robots Actuated by Cables with Handy and Neat Interface Software”, Cable-Driven Parallel Robots, Mech. and Mach. Sci., Vol. 32, pp. 293-305, 2015
[C+04] Q. Chen et al. An integrated two-level self-calibration method for cable-driven manipulator. IEEE Trans. on Robotics, 10(2):380–391, 2004.
[CM13] M. Carricato and J-P. Merlet. Stability analysis of underconstrained cable-driven parallel robots. IEEE Trans. on Robotics, 29(1):288–296, 2013.
[D+11] T. Dallej et al. Towards vision-based control of cable-driven parallel robots. IROS, San-Francisco, September, 25-30, 2011.
[ELTO16] J. Eden, D. Lau, Y. Tan, D. Oetomo, “Available Acceleration Set for the Study of Motion Capabilities for Cable-Driven Robots”, Mechanisms and Machine Theory, Vol. 105, pp. 320-336, 2016.
[FCG14] Fortin-Coté, A., Cardou, P., Gosselin, C.: An admittance control scheme for haptic interfaces based on cable-driven parallel mechanisms. In: Proc. of the IEEE Int. Conf. on Robotics and Automation (ICRA 2014), pp. 819–925. Hong Kong (2014)
[FM05] H. Fang and J-P. Merlet. Multi-criteria optimal design of parallel manipulators based on interval analysis. Mechanism and Machine Theory, 40(2):151–171, February 2005.
[Gag16] L. Gagliardini. Discrete reconfiguration of cable-driven parallel robots. PhD thesis, Ecole Centrale de Nantes, Nantes, September, 19, 2016.
[GCGG16] Gagliardini, L., Caro, S., Gouttefarde, M., Girin, A.: Discrete reconfiguration planning for cable-driven parallel robots. Mechanism and Machine Theory 100, 313–337 (2016)
[GNB14] M. Gouttefarde et al. Kinetostatics analysis of cable-driven parallel robots with consideration of sagging and pulleys. ARK, 2014.
[H+15] A. Hernandez et al. Design optimization of a cable-based parallel tracking system by using evolutionary algorithms. Robotica, 33(3):599–610, March 2015.
[HC04] Holland, C., Cannon, D.: Cable array robot for material handling (2004)
[HK09] Hassan, M.,Khajepour, A.: Analysis of large-workspace cable-actuated manipulator for warehousing applications. In: Proc. of the ASME Int. Design Eng. Tech. Conf. & Comput. and Inform. in Eng. Conf. (IDETC/CIE 2009), pp. 45–53. San Diego, CA (2009)
[I+12] J-B. Izard et al. A reconfigurable robot for cable-driven parallel robotic research and industrial scenario proofing. In 1st Int. Conf. on cable-driven parallel robots (CableCon), pages 135–148,Stuttgart, September, 3-4, 2012.
[JBK07] B.L. Jordan, M.A. Batalin, and W.J. Kaiser. NIMS RD: a rapidly deployable cable based robot. ICRA, pages 144–150, Roma, April, 10-14, 2007.
[JS03] S.A. Joshi and A. Surianarayan. Calibration of a 6-dof cable robot using two inclinometers. In Performance Metrics for Intelligent Systems, Gaithersburg, September, 16-18, 2003.
[KKW00] Kamawura, S., Kino, H., Won, C.: High-speed manipulation by using parallel wire-driven robots. Robotica 18(1), 13–21 (2000)
[K+06] K. Kozak et al. Static analysis of cable-driven manipulators with non-negligible cable mass. IEEE Trans. on Robotics, 22(3):425–433, June 2006.
[KB10] M.H. Korayem and M. Bamdad. Stiffness modeling and stability analysis of cable-suspended manipulators with elastic cable for maximum load determination. Kuwait J. Sci. Eng., 37(1b):181–201,2010.
[KKP15] W. Krauss, M. Kessler, and A. Pott. Pulley friction compensation for winch-integrated cable force measurement and verification on a cable-driven parallel robot. In IEEE Int. Conf. on Robotics andAutomation, pages 1627–1632, Seattle, May, 26-30, 2015.
[M06] J.P. Merlet, “Jacobian, Manipulability, Condition Number, and Accuracy of Parallel Robots”, ASME J. Mech. Des., Vol. 128, No. 1, 199–206, 2006.
[M+14] P. Miermeister et al. An elastic cable model for cable-driven parallel robots including hysteresis effect. In 2nd Int. Conf. on cable-driven parallel robots (CableCon), pages 17–28, Duisburg, August,24-27, 2014.
[MD10] Merlet, J.P., Daney, D.: A portable, modular parallel wire crane for rescue operations. ICRA 2010, pp. 2834–2839. Anchorage, AK (2010)
[Mer16a] J-P. Merlet. A new generic approach for the inverse kinematics of cable-driven parallel robot with 6 deformable cables. In ARK, 27-30, 2016.
[Mer16b] J-P. Merlet. A generic numerical continuation scheme for solving the direct kinematics of cable-driven parallel robot with deformable cables. In IEEE Int. Conf. on Intelligent Robots and Systems (IROS), Daejeon, October, 9-14, 2016.
[Mer17] J-P. Merlet. Simulation of discrete-time controlled cable-driven parallel robots on a trajectory. IEEE Trans. on Robotics, 33(3):675–688, 2017.
[MOr13] A. MOradi. Stiffness analysis of cable-driven parallel robot. PhD thesis, Queen's University,Kingston, April 2013.
[MP12] P. Miermeister and A. Pott. Auto calibration method for cable-driven parallel robot using force sensors. In ARK, Innsbruck, June, 25-28, 2012.
[N+13] D.Q. Nguyen et al. On the simplification of cable model in static analysis of large dimension cable-driven parallel robots. In IEEE Int. Conf. on Intelligent Robots and Systems (IROS), pages 928–934, Tokyo, November, 3-7, 2013.
[NGCP14] Nguyen, D.Q., Gouttefarde, M., Company, O., Pierrot, F.: On the analysis of large-dimension reconfigurable suspended cable-driven parallel robots. In: Proc. of the IEEE Int. Conf. on Robotics and Automation (ICRA 2014), pp. 5728–5735. Hong Kong (2014)
[PMV10] Pott, A., Meyer, C., Verl, A.: Large-scale assembly of solar power plants with parallel cable robots. In: Proc. of the Int. Symp. on Robotics and 6th German Conf. on Robotics (ISR/ROBOTIK 2010), pp. 1–6. Munich, Germany (2010)
[Pot12] A. Pott. Influence of pulley kinematics on cable-driven parallel robots. In ARK, pages 197–204, Innsbruck, June, 25-28, 2012.
[R+04] A.T. Riechel et al. Concept paper: cable-driven robots for use in hazardous environments. Topical Meeting on Robotics in hazardous environments, Gainesville, 2004.
[R+09] N. Riehl et al. Effects of non-negligible cable mass on the static behavior of large workspace cable-driven parallel mechanisms. In IEEE Int. Conf. on Robotics and Automation, pages 2193–2198,Kobe, May, 14-16, 2009.
[R+10a] N. Riehl et al. On the static workspace of large dimension cable-suspended robots with non negligible cable mass. In 34th Annual Mechanisms and Robotics Conference, Montréal, August, 15-18,2010.
[R+10b] N. Riehl et al. On the determination of cable characteristics for large dimension cable-driven parallel mechanisms. In IEEE Int. Conf. on Robotics and Automation, pages 4709–4714, Anchorage, May,3-8, 2010.
[SC82] J.K. Salisbury, J.J. Craig, “Articulated Hands: Force Control and Kinematic Issues”, Int. Journal of Robotics Research, Vol. 1, No. 4, pp. 4–17, 1982.
[SWI17] D. Sridhar and R.L. Williams II. Kinematics and statics including cable sag for large cable suspended robots. Global Journal of Researches in Engineering: H Robotics & Nano-Tec, 17(1), 2017.
[S+09] M. Such et al. An approach based on the catenary equation to deal with static analysis of three dimensional cable structures. Engineering structures, 31(9):2162–2170, 2009.
[Y83] T. Yoshikawa, “Analysis and Control of Robot Manipulators with Redundancy”, First Int. Symposium on Robotics Research, pp. 735–747, 1983.
[YTWH10] Yao, R., Tang, X., Wang, J., Huang, P.: Dimensional optimization design for the four-cable driven parallel manipulator in FAST. IEEE/ASME Trans. on Mechatronics 15(6), 932–941 (2010)

%%%
Prerequisites:

The candidate must hold an engineering degree and/or a master's degree in advanced mechanical engineering or robotics. The position requires a good knowledge of:

1) Mechanical design
2) Modeling of polyarticulated and flexible systems
3) Robotic system control

as well as

- good oral and written communication skills (writing articles, presenting at meetings or congresses, interactions)
- a very good level in English writing and speaking (TOEIC level > 900)
- an ability to work in a team


View or Apply

Similar Positions