Smart Robotic Assistants for Small Volume Manufacturing Tasks

Size: px

Start display at page:

Download "Smart Robotic Assistants for Small Volume Manufacturing Tasks"

Winifred Charles
6 years ago
Views:

1 Smart Robotic Assistants for Small Volume Manufacturing Tasks Satyandra K. Gupta Director, Center for Advanced Manufacturing Smith International Professor Aerospace and Mechanical Engineering Department Viterbi School of Engineering University of Southern California

2 Examples of Applications for Traditional Industrial Robots ASSEMBLY WELDING (Image Source: (Image Source: LOADING AND UNLOADING (Image Source: PAINTING (Image Source:

3 Characteristics of Traditional Industrial Robots Cannot share space with humans Need human experts to program robots Able to perform simple repetitive tasks Require significant lead time to develop custom hardware Installed at fixed locations Require high utilization rate to justify return on investment Characteristics of traditional industrial robots make it difficult to use them in small volume manufacturing

4 Why use robots in small volume manufacturing applications? Tasks that humans cannot do High speed composite layup to improve part quality Tasks that human should not do Handling of heavy parts Tasks that humans don t want to do Sanding and polishing, support structure removal from 3D printed metal parts Small volume manufacturing can benefit tremendously from use of industrial robots!

5 Current Situation New capabilities are removing traditional barriers Industrial robot costs have come down It is now possible to consider using robots on small volume manufacturing applications

6 Our Goal Develop physics-aware decision making methodologies and computational tools to enable use of robots on non-repetitive manufacturing tasks

7 Examples of Small Volume Manufacturing Applications being developed at USC Center for Advanced Manufacturing

performed on complex geometries require constant monitoring and adjustment Source: http://www.geeksoncars.

8 Robotic Finishing: Motivation Finishing tasks can take up to 25% of manufacturing costs Examples: cleaning, sanding, polishing, de-burring, support material removal for 3D printing, rust removal Finishing operation performed on complex geometries require constant monitoring and adjustment Source: A large number of finishing tasks are currently performed manually!

tool trajectory parameters using active learning Combination of data-driven and model-driven

9 Robotic Finishing: Approach On-line Planning Setup planning to determine optimal number of setups Grasp location planning Generation of tool trajectories for each setup On-line Learning Selection of tool trajectory parameters using active learning Combination of data-driven and model-driven approach to learn cleaning process model Task Safety Impedance control and on-line perception for task safety

10 Robotic Finishing: Videos VWiGuaoxL0sGNw-PNi8r3h2tK6zz (Advanced Robotic Finishing KUKA Innovation Award 2017 Finalist Spotlight) (Robotic Finishing Demo with User Interface) (Robotic Finishing Motivation) (Robotic Finishing Experiments Compilation) (Robotic Cleaning) (Robotic Cleaning of Compliant Objects)

11 Robotic Bin Picking for Complex Parts: Motivation Currently robotic bin picking is used only for simple parts Robotic bin picking is not used in mixed bins or parts that have tendency to tangle

Robotic Bin Picking for Complex Parts: Approach Automated perception to estimate part location in the bin and associated uncertainty Extraction planning to remove the part

12 Robotic Bin Picking for Complex Parts: Approach Automated perception to estimate part location in the bin and associated uncertainty Extraction planning to remove the part from the bin without causing the part to be tangled with other parts in the bin If the task completion confidence is low, then the system seeks help from the human operator

13 Robotic Bin Picking for Complex Parts: Video (RoboSAM: Robotic Smart Assistant for Manufacturing)

Constrained by deposition orientation and geometric irregularity

14 Robotic Additive Manufacturing: Motivation Traditional additive manufacturing uses planar layers 3D object converted to 2D layers Constrained by deposition orientation and geometric irregularity Building curved geometries takes a long time Strength is compromised

support material Reduces cost, time, and material waste Ability to exploit

15 Robotic Additive Manufacturing: Approach Robots enable deposition of non-planar layers Minimize number of layers Improve strength Reduce support material Reduces cost, time, and material waste Ability to exploit composite material Capability of material deposition on irregular surfaces and platforms

16 Robotic Additive Manufacturing: Video (Additive Manufacturing with Non-planar Layers)

Collaborative Assembly: Motivation Robots are used only on high volume assembly applications Low volume applications will require collaboration between humans and robots

17 Collaborative Assembly: Motivation Robots are used only on high volume assembly applications Low volume applications will require collaboration between humans and robots Challenging tasks will need to be assigned to humans Robots should be able to seek help from humans to handle contingencies Source: GE, Photo: JIM R. BOUNDS, BLOOMBERG NEWS

the task progress during the execution Real-time replanning algorithms

18 Collaborative Assembly: Approach Task planning and resource allocation to explicitly account for managing contingencies Real-time monitoring of the task progress during the execution Real-time replanning algorithms to handle contingencies Human robot collaboration to deal with contingencies

19 Collaborative Assembly: Video (Robotic Assistants to Support Complex Assembly Operations in Small Production Volumes) (Assembly automation using smart robotic assistant) (A Flexible Hybrid Cell for Low Production Volume Assembly)

execution to prevent defects in the part http://d2n4wb9orp1vta.cloudfront.

20 Robotic Composite Layup: Motivation Automated tape layup (ATL) machines can only handle simple geometries Complex layup operations will require collaboration among multiple robots and humans Task monitoring needs to be integrated with task execution to prevent defects in the part n/cms/sb09_compositesthematerials_h.jpg dxt5xhaejru/tq_vlw1m3ri/aaaaaaaabiw/8dgwvcw3ouu /s1600/hand_lay_up_01.jpg

Multi-arm manipulation of deformable sheets Trajectory

21 Robotic Composite Layup: Approach Design of end effectors for material dispensing and layup on complex molds Multi-arm manipulation of deformable sheets Trajectory planning and refinement Efficient and safe human robot collaboration

22 Robotic Composite Layup: Video (Collaborative Robotic Cell for Composite Sheet Layup) (Hybrid Cells for Multi-Layer Prepreg Composite Sheet Layup)

to model Interactions among component technologies We have demonstrated feasibility of smart robotic assistants in new

23 Conclusions Recent advances in the field of industrial robots are adding new capabilities at an unprecedented rate Physics-aware decision making is a key to realizing smart robotic assistants Application context is very important Need to model Interactions among component technologies We have demonstrated feasibility of smart robotic assistants in new manufacturing applications Finishing Bin Picking for Complex Parts Collaborative Assembly Composite Layup Additive Manufacturing

24 References Alsharhan, A. T., Centea, T., Gupta, S. K., Enhancing mechanical properties of thin-walled structures using non-planar extrusion based additive manufacturing. ASME Manufacturing Science and Engineering Conference, Kabir, A. M., Langsfeld, J. D., Zhaung, C., Kaipa, K. N., and Gupta, S. K. A systematic approach for minimizing physical experiments to identify optimal trajectory parameters for robots. IEEE International Conference on Robotics and Automation, Kabir, A. M. and Kaipa, K. N. and Marvel, J. and Gupta, S. K. Automated Planning for Robotic Cleaning using Multiple Setups and Oscillatory Tool Motions. IEEE Transactions on Automation Science and Engineering, Kabir, A. M., Langsfeld, J. D., Zhuang, C., Kaipa, K. N., Gupta, S. K. Automated Leaning of Operation Parameters for Robotic Cleaning by Mechanical Scrubbing. ASME Manufacturing Science and Engineering Conference, Kabir, A. M., Langsfeld, J. D., Shriyam, S., Rachakonda, V., Zhaung, C., Kaipa, K. N., Marvel, J., and Gupta, S. K. Planning Algorithms for Multi-Setup Multi-Pass Robotic Cleaning with Oscillatory Moving Tools. IEEE International Conference on Automation Science and Engineering, 2016.

25 References (Cont.) Kaipa, K. N., Kankanhalli-Nagendra, A. S., Kumbla, N. B., Shriyam, S., Thevendria-Karthic, S. S., Marvel, J., Gupta, S. K. Addressing Perception Uncertainty Induced Failure Modes in Robotic Bin-Picking, Robotics and Computer-Integrated Manufacturing, Kaipa, K. N., Shriyam, S., Kumbla, N. B., Gupta, S. K., Resolving Occlusions Through Simple Motions in Robotic Bin-Picking, ASME Manufacturing Science and Engineering Conference, Kaipa, K. N., Shriyam, S., Kumbla, N. B., Gupta, S. K. Automated Plan Generation for Robotic Singulation from Mixed Bins. IROS Workshop on Task Planning for Intelligent Robots in Service and Manufacturing, Kaipa, K. N., Kumbla, N. B., Gupta, S. K. Characterizing Performance of Sensorless Fine Positioning Moves in the Presence of Grasping Position Uncertainty, IROS Workshop on Task Planning for Intelligent Robots in Service and Manufacturing, Kaipa, K. N., Thevendria-Karthic, S. S., Shriyam, S., Kabir, A. M., Langsfeld, J. D., & Gupta, S. K., Resolving automated perception system failures in bin-picking tasks using assistance from remote human operators. IEEE International Conference on Automation Science and Engineering, 2015 Kumbla, N. B., Thakar, S., Kaipa, K. N., Marvel, J., Gupta, S. K., Simulation Based On-Line Evaluation of Singulation Plans to Handle Perception Uncertainty in Robotic Bin Picking, ASME Manufacturing Science and Engineering Conference, 2017.

26 References (Cont.) Langsfeld, J. D., Kabir, A. M., Kaipa, K. N., Gupta, S. K. Online Learning of Part Deformation Models for Robotic Cleaning. ASME Manufacturing Science and Engineering Conference, Langsfeld, J. D., Kabir, A. M., Kaipa, K. N., Gupta, S. K. Robotic Bimanual Cleaning of Deformable Objects with Online Learning of Part and Tool Models. IEEE International Conference on Automation Science and Engineering, Morato, C. W., Kaipa, K. N., Liu, J., & Gupta, S. K. A framework for hybrid cells that support safe and efficient human-robot collaboration in assembly operations. ASME Computers and Information in Engineering Conference, August Morato, C., Kaipa, K., Zhao, B., & Gupta, S. K. Safe human robot interaction by using exteroceptive sensing based human modeling. ASME Computers and Information in Engineering Conference, August 2013

27 References (Cont.) Morato, C., Kaipa, K. N., Zhao, B., Gupta, S. K. Toward safe human robot collaboration by using multiple kinects based real-time human tracking. Journal of Computing and Information Science in Engineering, 14(1), , Morato, C., Kaipa, K. N., & Gupta, S. K. Improving assembly precedence constraint generation by utilizing motion planning and part interaction clusters. Computer-Aided Design, 45(11), , 2013.

1. PERSONAL INFORMATION A. RESEARCH INTERESTS. Computer Aided Design, Manufacturing Automation, and Robotics

1. PERSONAL INFORMATION A. RESEARCH INTERESTS. Computer Aided Design, Manufacturing Automation, and Robotics Curriculum Vitae of Dr. Satyandra K. Gupta 3650 McClintock Ave., OHE 430G Aerospace and Mechanical Engineering Department University of Southern California, Los Angeles, California 90089-1453 Phone: (213)