Machine Vision for Collaborative Robot Applications. David L. Dechow FANUC America Corporation

Transcription

1 Machine Vision for Collaborative Robot Applications David L. Dechow FANUC America Corporation

2 Topics Overview of collaborative robot technologies The roles for machine vision It s still machine vision - Unique challenges Applications and future trends A disclaimer

3 Collaborative Robot Overview Defining the collaborative robot The COBOT landscape

4 Collaborative Robot Overview Safety considerations ANSI/RIA specifications ISO/TS How they might be implemented

5 Collaborative Robot Overview Safety considerations

6 Collaborative Robot Overview Safety Considerations Safety Rated Monitored Stop Safe I/O input from safety rated device (scanner or light screen) Input to DCS enables speed limit of 0 mm/s or 0 deg/s Robot program can remain in cycle, no e-stop issues Once the operator leaves the protected area, production can resume with out any delay for faults to clear or servo power to be re-enabled

7 Collaborative Robot Overview Safety Considerations Hand Guiding

8 Collaborative Robot Overview Safety Considerations Speed and Separation Monitoring Completely fenceless operation Warning Zone #1: Speed Override to 250mm/sec Warning Zone #2: Program Pause Safety Zone: DCS Zero Speed Zone Operating Space Operating Space 18 + Stopping Distance Semi-Fenceless Concept using DCS Position Check and a Safety Area Scanner 18 + Stopping Distance

9 Collaborative Robot Overview Safety Considerations Power and Force Limiting Sensitivity varies Pinch points must be considered Speed and payloads impact the safety requiremets Push J2 Force and power limiting can be achieved by monitoring and stopping robot based on force detected between a robot and operator Force and power limiting = Contact Stop Retract Stop Push J1

10 Collaborative Robot Overview Safety Considerations Integrators and/or Users MUST do a risk assessment Having a collaborative robot in the cell does not make the entire cell safe Design must mitigate hazards in other cell components including the part and tooling Possible contact between a person and robot must not result in injury What are appropriate values for force, when the operator and robot collide, without causing pain or injury The international standards groups have been looking at this problem and have preliminary guidance for what is appropriate based on medical studies This needs to be balanced with the tasks and possible contacts within a cell ISO TS15066 helps address these questions

11 Collaborative Robot Overview Applications and limitations Many uses Part installation and assembly Part delivery Machine tending Palletizing/packing/depalletizing Meeting speed and payload expectations

12 The Roles For Machine Vision The two sides of collaboration Working autonomously or together Safe robots working in the vicinity of human workers Robot and human collaboration can we see each other?

13 The Roles For Machine Vision More-flexibile automation Human-like performance means human-like environments True collaboration is elusive with respect to machine vision technology VGR for the collaborative robot standard techniques and perhaps much more

14 It s Still Machine Vision It s still machine vision Similar methods, new challenges Ease of use at highest demand Complex application requirements impact machine vision implementations Support tools can ease robot programming Broader recognition capabilities are at a premium 3D is more in demand

15 It s Still Machine Vision Some VGR basics 2D guidance finding and delivering the object

16 It s Still Machine Vision Some VGR basics 3D guidance now a requirement

17 It s Still Machine Vision Challenges old and new Lighting in the collaborative environment The human interference factor Handling random parts

18 It s Still Machine Vision Lighting in the collaborative environment Homogeneous illumination over a large area? Handling the human Field of View

19 It s Still Machine Vision The human interference factor Machine vision for finding the person not the part Finding parts when the human is in the way Seeing and avoiding interference

20 It s Still Machine Vision Handling random parts Applying existing 3D guidance techniques for collaborative applications

21 Applications And Future Trends Application videos

22 Applications And Future Trends Application videos

23 Applications And Future Trends What s next The market is justifiably constrained by safety demands More machine vision capability will drive future applications Technology is almost ready for prime time

24 Contact Information David L. Dechow Staff Engineer-Intelligent Robotics/Machine Vision FANUC America Corporation 3900 W. Hamlin Road Rochester Hills, MI (248)