Climbing Feet for RiSE June 14, 2004 Miguel Piedrahita BDML

Overview of Future Climbing Foot Development Long-term goals: –Sufficient shear and normal forces for dynamic vertical climbing, on a variety of surfaces –Effective attachment and detachment –Integration of dry adhesives with spines and perhaps other mechanisms –Integrated sensing Short-term needs: –We need to demonstrate effective climbing with the RiSE platform for key dates (August, October)!

What do we need for success in Aug/Oct? Video which shows progress towards vertical climbing on a variety of surfaces Show clear integration path of innovations from demonstration platform

58.0 deg Based on location of CM and rear feet, we can stand on 58 deg slope without adhesion. Could we climb on a slightly flatter slope? Still need lots of shear force! Is that good enough for Aug/Oct? Do we need adhesion?

Surfaces Surface Type “Soft” (penetrable) “Rough” (impenetrable) “Smooth” (impenetrable) ExamplesCork board, soft woods Concrete, hard woods Glass, metal, painted wall Good climbing strategies Claws, spines with little compliance Claws, spines with high compliance Dry adhesives, PSA, friction pads Original M18 goals 90 deg60 deg30 deg

Specialized Feet for each surface? Easiest way to achieve good climbing on these three types of surfaces is to design a separate foot for each surface! –Soft surface: Larger, stiffer claws and spines –Rough surface: Delicate, individually sprung claws and spines –Smooth surface: PSA and friction pads Will RiSE researchers and DARPA buy into this approach? Not a very elegant solution, will lead to integration challenges in the future.

Single Climbing Foot for all Surfaces? Requirements: –Foot-level Compliance High yaw compliance to deal with changing foot angle during stroke due to 4-bar kinematics Some pitch & roll compliance to conform to surface –Effectively combine several adhesion/shear mechanisms OR –Develop a single adhesion/shear mechanism which will work on all surfaces

Combine several mechanisms? Retractable spines/claws –Push/pull cables, SMA Spines/claws with changing compliance Full-foot pivoting to engage different foot surfaces –R/C servo or SMA actuation –Requires smart actuation (surface-dependent) Surface with firm spines Surface with compliant spines & adhesive pad

Single Mechanism for all surfaces? Clever sprung claw design, like Moto’s hummingbird foot –Maximize penetration on soft surfaces –Minimize push-off force on smooth surfaces

What needs to be done Decide on mechanism(s) to include in foot Consider front/rear foot specialization! Test current foot designs on test track: –Measure maximum adhesion and shear force from each foot design –Do different feet have different optimum trajectories? FEA and hand calcs to determine good geometries for polyurethane or dry adhesive pads FEA and hand calcs to develop flexures