Automatic selection of ergonomic indicators for the design of collaborative robots: a virtual-human in the loop approach P. Maurice, P. Schlehuber, Y. Measson, V. Padois, P. Bidaud F R O M R E S E A R C H T O I N D U S T R Y 2014 IEEE-RAS International Conference on Humanoid Robots November 19, 2014
Introduction Work-related musculoskeletal disorders: A major health problem Statistics [Schneider and Irastorza, 2010] Main biomechanical risk factors ◮ Affect over 35 % of workers in Europe ◮ Extreme postures ◮ Represent the 1st occupational disease ◮ Considerable efforts ◮ Increase by 15 % per year ◮ Static work ◮ High frequency of the gestures ◮ Cost about $50B a year in the US Tension neck Rotator cuff syndrome tendinitis Epicondylitis Carpal tunnel Low back pain syndrome Bursitis Achilles tendinitis Pauline MAURICE Selection of ergonomic indicators for collaborative robotics using a DHM Humanoids 2014 1 / 11
Introduction Collaborative robotics: A physical assistance for complex tasks Robot [Colgate et al. , 2003] Human ◮ Weight compensation ◮ Technical expertise ◮ Strength amplification ◮ Adaptability ◮ Guidance via virtual paths ◮ Decision Co-manipulation of objects or tools Pauline MAURICE Selection of ergonomic indicators for collaborative robotics using a DHM Humanoids 2014 2 / 11
Introduction Limitations of DHM based ergonomic assessments Macroscopic models Biomechanical models ◮ Products: Delmia, Jack 1 , ◮ Products: OpenSim 3 , Anybody 4 , Sammie 1 , 3DSSPP 2 . . . LifeMOD, Santos . . . ◮ Ergonomic assessments: ◮ Ergonomic assessments: RULA 5 , OWAS 5 , Snook tables 6 , Joint force, Muscle force, Tendon NIOSH 7 , Low-back analysis . . . length . . . Rough or Task-specific Numerous criteria One global criterion Accurate and Generic 1 [Delleman et al. , 2004], 2 [Chaffin et al. , 2006], 3 [Delp et al. , 2007], 4 [Damsgaard et al. , 2006], 5 [Li and Buckle, 1999], 6 [Snook and Ciriello, 1991], 7 [Waters et al ., 1993] Pauline MAURICE Selection of ergonomic indicators for collaborative robotics using a DHM Humanoids 2014 3 / 11
Introduction Limitations of DHM based ergonomic assessments Macroscopic models Biomechanical models ◮ Products: Delmia, Jack 1 , ◮ Products: OpenSim 3 , Anybody 4 , Sammie 1 , 3DSSPP 2 . . . LifeMOD, Santos . . . ◮ Ergonomic assessments: ◮ Ergonomic assessments: RULA 5 , OWAS 5 , Snook tables 6 , Joint force, Muscle force, Tendon NIOSH 7 , Low-back analysis . . . length . . . Rough or Task-specific Numerous criteria One global criterion One global criterion Accurate and Generic Accurate and Generic Selection of relevant ergonomic indicators ◮ Dedicated to the comparison of collaborative robots ◮ Dependent on task features ◮ Independent from robot design ◮ Automatic DHM-based process 1 [Delleman et al. , 2004], 2 [Chaffin et al. , 2006], 3 [Delp et al. , 2007], 4 [Damsgaard et al. , 2006], 5 [Li and Buckle, 1999], 6 [Snook and Ciriello, 1991], 7 [Waters et al ., 1993] Pauline MAURICE Selection of ergonomic indicators for collaborative robotics using a DHM Humanoids 2014 3 / 11
Method Introduction 1 Method 2 Results 3 Conclusion 4 Pauline MAURICE Selection of ergonomic indicators for collaborative robotics using a DHM Humanoids 2014 3 / 11
Method Indicators relevance: Differentiating various ways of performing a task Human and robot parameters Selection ... Parameters set #1 Parameters set #N Dynamic simulation Manikin Task Robot controller controller description Force amplification LQP ... Indicators set #1 Indicators set #N Analysis Relevant ergonomic indicators Pauline MAURICE Selection of ergonomic indicators for collaborative robotics using a DHM Humanoids 2014 4 / 11
Method Parameters selection: Creating a variety of situations Parameters ◮ Amplification coefficient ◮ Robot mass ◮ Upper body joint limits ◮ Pelvis position ◮ Upper body reference posture ◮ Upper body tasks weights ◮ Step length ◮ Human size ◮ Human body mass index F rob = α F vh K F rob τ rob = α J T F vh + g ( q ) M Exploration B Collaborative robot Abstraction of the robot Fourier amplitude sensitivity testing (FAST) [Saltelli et al. , 1999] Pauline MAURICE Selection of ergonomic indicators for collaborative robotics using a DHM Humanoids 2014 5 / 11
Method A dynamic DHM for indicators calculation Tasks Constraints ◮ Balance: ZMP preview control ◮ Dynamical model equation ◮ Hands trajectories and forces ◮ Joint limits ◮ Whole body posture ◮ Joint torques saturation ◮ Torques minimization ◮ Non sliding contacts Optimization [Salini et al. , 2011] Linear Quadratic Programming with weighting strategy aa � Joint torques � Ergonomic Manikin state bb Contact forces indicators ”Torques” Dynamic simulation ”Robot” controller XDE framework (CEA-LIST) Force amplification Robot state Interaction force Pauline MAURICE Selection of ergonomic indicators for collaborative robotics using a DHM Humanoids 2014 6 / 11
Method Ergonomic indicators selection: A variance-based analysis Indicator 1 ... Indicator N Local indicators ◮ Positions ◮ Back � � ◮ Velocities ... I 1 ( t ) dt I N ( t ) dt ◮ Right arm ◮ Accelerations task task ◮ Left arm ◮ Torques ◮ Legs ◮ Power ... Scaling Scaling Global indicators ◮ Kinetic energy ... Variance Variance ◮ Force transmission ratio ◮ Velocity transmission ratio ◮ Balance robustness ◮ Dynamic balance Scree test: elbow criterion Discriminating indicators Pauline MAURICE Selection of ergonomic indicators for collaborative robotics using a DHM Humanoids 2014 7 / 11
Method Ergonomic indicators selection: A variance-based analysis Indicator 1 ... Indicator N Scaling value ◮ mean( I i / I ref ) = 1 i � � ◮ variance( I i / I ref ... ) � = 1 I 1 ( t ) dt I N ( t ) dt i task task I m , n � � i m ∈ T n ∈ P I ref = i N T N P ... Scaling Scaling Scaling Scaling T : tasks, P : parameters sets ... Variance Variance Scree test: elbow criterion Discriminating indicators Pauline MAURICE Selection of ergonomic indicators for collaborative robotics using a DHM Humanoids 2014 7 / 11
Method Ergonomic indicators selection: A variance-based analysis Indicator 1 ... Indicator N Scaling value ◮ mean( I i / I ref ) = 1 i � � ◮ variance( I i / I ref ... ) � = 1 I 1 ( t ) dt I N ( t ) dt i task task I m , n � � i m ∈ T n ∈ P I ref = i N T N P ... Scaling Scaling T : tasks, P : parameters sets Scree plot ... Variance Variance Variances Scree test: elbow criterion Scree test: elbow criterion elbow selected not selected Discriminating indicators I1 I2 I3 I4 I5 I6 I7 Indicators Pauline MAURICE Selection of ergonomic indicators for collaborative robotics using a DHM Humanoids 2014 7 / 11
Results Introduction 1 Method 2 Results 3 Conclusion 4 Pauline MAURICE Selection of ergonomic indicators for collaborative robotics using a DHM Humanoids 2014 7 / 11
Results Application to various tasks: Selected indicators Walk Walk Reach Reach Fast traj. Fast traj. Push Push Bend and Bend and sideways sideways 35 cm 35 cm tracking tracking 100 N 100 N Carry 3 kg Carry 3 kg Kinetic energy Velocity Transmission Ratio Force Transmission Ratio Dynamic balance Balance robustness Legs torque Legs power Legs acceleration Legs velocity Legs position ◮ 3 to 8 indicators selected Left arm torque Left arm power out of 29 Left arm acceleration Left arm velocity ◮ Variance information loss Left arm position < 20 % Right arm torque Right arm power Right arm acceleration Right arm velocity Right arm position Back torque Back power Back acceleration Back velocity Back position 81 % 80 % 86 % 81 % 82 % Pauline MAURICE Selection of ergonomic indicators for collaborative robotics using a DHM Humanoids 2014 8 / 11
Results Application to various tasks: Illustration of the parameters effects Pauline MAURICE Selection of ergonomic indicators for collaborative robotics using a DHM Humanoids 2014 9 / 11
Conclusion Introduction 1 Method 2 Results 3 Conclusion 4 Pauline MAURICE Selection of ergonomic indicators for collaborative robotics using a DHM Humanoids 2014 9 / 11
Conclusion Conclusion: Automatic selection of ergonomic indicators Constraints Method ◮ Dedicated to collaborative ◮ Parameters: Varying human and robotics robot (abstraction) features ◮ Independent from robot design ◮ Dynamic simulation: DHM with LQP based controller ◮ Dependent on task features ◮ Indicators: Variance-based ◮ Automatic analysis of mutliple biomechanical ◮ Differentiate various ways of quantities performing a task Results ◮ Physically consistent selection ◮ 6 relevant indicators on average ◮ > 80 % information remains Pauline MAURICE Selection of ergonomic indicators for collaborative robotics using a DHM Humanoids 2014 10 / 11
Recommend
More recommend
