Control of biped locomotion inspired from inspired from walking in - PowerPoint PPT Presentation

Control of biped locomotion Control of biped locomotion inspired from inspired from walking in walking in monkeys monkeys Charles-Henry Houdemer Introduction Charles-Henry Houdemer Insights on BMIs Previous work Goal of the project Challenges Biorob Long term goals Ecole polytechnique fédérale de Lausanne Results Webots model LSRO Numerical tests École polytechnique fédérale de Lausanne CPG model Results Future 18 June 2010 / Final Presentation improvements References Supervisors : Sarah Dégallier, Jesse Van den Kieboom, Solaiman Shokur, Auke Jan Ijspeert

Outline Control of biped Introduction 1 locomotion inspired from Insights on BMIs walking in monkeys Previous work Charles-Henry Goal of the project Houdemer Challenges Introduction Long term goals Insights on BMIs Previous work Goal of the project Challenges Results 2 Long term goals Webots model Results Webots model Numerical tests Numerical tests CPG model CPG model Results Results Future improvements References Future improvements 3

Insights on BMIs Control of biped locomotion inspired from walking in monkeys Three different forms : Charles-Henry Invasive Houdemer Partially-invasive Introduction Non-invasive Insights on BMIs Previous work Two types : Goal of the project Challenges Motor BMI Long term goals Sensory BMI Results Webots model Transform data either from or to the brain by predicting Numerical tests CPG model and interpreting correct behaviour Results Future improvements References

Insights on BMIs Control of biped locomotion inspired from walking in monkeys Three different forms : Charles-Henry Invasive Houdemer Partially-invasive Introduction Non-invasive Insights on BMIs Previous work Two types : Goal of the project Challenges Motor BMI Long term goals Sensory BMI Results Webots model Transform data either from or to the brain by predicting Numerical tests CPG model and interpreting correct behaviour Results Future improvements References

Insights on BMIs Control of biped locomotion inspired from walking in monkeys Three different forms : Charles-Henry Invasive Houdemer Partially-invasive Introduction Non-invasive Insights on BMIs Previous work Two types : Goal of the project Challenges Motor BMI Long term goals Sensory BMI Results Webots model Transform data either from or to the brain by predicting Numerical tests CPG model and interpreting correct behaviour Results Future improvements References

Insights on BMIs Control of biped locomotion inspired from walking in monkeys Three different forms : Charles-Henry Invasive Houdemer Partially-invasive Introduction Non-invasive Insights on BMIs Previous work Two types : Goal of the project Challenges Motor BMI Long term goals Sensory BMI Results Webots model Transform data either from or to the brain by predicting Numerical tests CPG model and interpreting correct behaviour Results Future improvements References

Insights on BMIs Control of biped locomotion inspired from walking in monkeys Three different forms : Charles-Henry Invasive Houdemer Partially-invasive Introduction Non-invasive Insights on BMIs Previous work Two types : Goal of the project Challenges Motor BMI Long term goals Sensory BMI Results Webots model Transform data either from or to the brain by predicting Numerical tests CPG model and interpreting correct behaviour Results Future improvements References

Previous work Control of biped locomotion inspired from walking in monkeys Charles-Henry Fitzsimmons’ work on monkeys 1 : Houdemer Model try to predict the position of each joint Introduction Insights on BMIs Quality of predictions is good for one direction but Previous work degrades for two directions Goal of the project Challenges Must switch between two models for forward and Long term goals Results backward Webots model Numerical tests CPG model Results Future improvements References

Experimental setup Control of biped locomotion inspired from walking in monkeys Charles-Henry Houdemer Introduction Insights on BMIs Previous work Goal of the project Challenges Long term goals Results Webots model Numerical tests CPG model Results Future improvements References

Achievements Control of biped locomotion inspired from walking in monkeys Charles-Henry Houdemer Introduction Insights on BMIs Previous work Goal of the project Challenges Long term goals Results Webots model Numerical tests CPG model Results Future improvements References

Goal of the project Control of biped locomotion inspired from Build a CPG with the data from Nicolelis Lab in Duke walking in monkeys University Charles-Henry Data consists in Cartesian positions of hip, knee and Houdemer ankle positions, speed of walking and indication if the Introduction foot is on the ground Insights on BMIs Previous work Reproduce the walking gait of a rhesus monkey with Goal of the project Challenges Long term goals this CPG recieving basic inputs from a BMI Results Generalize it if possible to : Webots model Numerical tests Different sessions; CPG model Results Different speeds; Future Different monkeys; improvements Both directions. References

Goal of the project Control of biped locomotion inspired from Build a CPG with the data from Nicolelis Lab in Duke walking in monkeys University Charles-Henry Data consists in Cartesian positions of hip, knee and Houdemer ankle positions, speed of walking and indication if the Introduction foot is on the ground Insights on BMIs Previous work Reproduce the walking gait of a rhesus monkey with Goal of the project Challenges Long term goals this CPG recieving basic inputs from a BMI Results Generalize it if possible to : Webots model Numerical tests Different sessions; CPG model Results Different speeds; Future Different monkeys; improvements Both directions. References

Goal of the project Control of biped locomotion inspired from Build a CPG with the data from Nicolelis Lab in Duke walking in monkeys University Charles-Henry Data consists in Cartesian positions of hip, knee and Houdemer ankle positions, speed of walking and indication if the Introduction foot is on the ground Insights on BMIs Previous work Reproduce the walking gait of a rhesus monkey with Goal of the project Challenges Long term goals this CPG recieving basic inputs from a BMI Results Generalize it if possible to : Webots model Numerical tests Different sessions; CPG model Results Different speeds; Future Different monkeys; improvements Both directions. References

Goal of the project Control of biped locomotion inspired from Build a CPG with the data from Nicolelis Lab in Duke walking in monkeys University Charles-Henry Data consists in Cartesian positions of hip, knee and Houdemer ankle positions, speed of walking and indication if the Introduction foot is on the ground Insights on BMIs Previous work Reproduce the walking gait of a rhesus monkey with Goal of the project Challenges Long term goals this CPG recieving basic inputs from a BMI Results Generalize it if possible to : Webots model Numerical tests Different sessions; CPG model Results Different speeds; Future Different monkeys; improvements Both directions. References

Challenges Control of biped locomotion inspired from walking in monkeys Charles-Henry Houdemer Monkeys are not naturally bipedal Monkeys’ behaviour is not always predictable Introduction Insights on BMIs Play with holding bar Previous work Goal of the project Make feet slide on the ground Challenges Long term goals Video Results Causes variability in measurements Webots model Numerical tests CPG model Results Future improvements References

Typical variations in the hip angles Control of biped locomotion inspired from walking in monkeys Charles-Henry Houdemer Introduction Insights on BMIs Previous work Goal of the project Challenges Long term goals Results Webots model Numerical tests CPG model Results Future improvements References

Long term goals Control of biped locomotion inspired from walking in monkeys Charles-Henry Houdemer Allow paraplegics to regain the control of their legs Introduction Allow for more robust predictions Insights on BMIs Previous work Less informations are needed Goal of the project Challenges Less neurons are needed Long term goals May lose more neurons before losing efficency Results Webots model Numerical tests CPG model Results Future improvements References

Long term goals Control of biped locomotion inspired from walking in monkeys Charles-Henry Houdemer Allow paraplegics to regain the control of their legs Introduction Allow for more robust predictions Insights on BMIs Previous work Less informations are needed Goal of the project Challenges Less neurons are needed Long term goals May lose more neurons before losing efficency Results Webots model Numerical tests CPG model Results Future improvements References