Design and Modeling of a Mobile Robot with an Optimal Obstacle-Climbing Mode The pen WHEEL Project Jean-Christophe FAUROUX Morgann FORLOROU Belhassen Chedli BOUZGARROU Frédéric CHAPELLE 1/33 LaMI / TIMS / IFMA / UBP,, Véhicules en Milieux Naturels
Summary 2. Existing mobile robots 3. Design workspace for mobile robots with climbing capacities 4. Structural design 5. Designing a climbing sequence 6. Conclusions and future work 2/33
Introduction Terrestrial vehicles Wheeled vehicles prevail (energetic efficiency?) Blocked on slope discontinuities of the ground Legs / Tracks regain interest for climbing Interface with the ground Crawler + multiples contacts, can cross obstacles and rough terrain - require high energy, moderate speed, complex control Leg + can cross obstacles and go fast on rough terrain - contact discontinuity, energy cost, stability control Wheel + fast on smooth surface, energy efficient - cannot climb obstacles or run on rough terrain Track + permanent stability, high traction - high friction energy loss, particularly during steering Mechanical architecture 3/33 No locomotion mode is perfect Improved wheeled architectures should be developed Engines distributed on the wheels Innovative architectures already exist (e.g. for spatial robots)
Existing mobile robots Robots with articulated frame Robots with axles and several modes of locomotion RobuROC 6 www.robosoft.fr Micro5 www.mit.edu/~ykuroda/ Lama www.laas.fr Nomad www.frc.ri.cmu.edu/projects/lorax 4/33 Hilos www.robot.jussieu.fr Robots with legs and several modes of locomotion Azimut www.gel.usherbrooke.ca/ laborius/ Workpartner / Hybtor www.automation.hut.fi
Existing mobile robots Robots with minimally actuated frame Crab http://www.asl.ethz.ch Shrimp http://www.asl.ethz.ch Robots with rocker-bogie and stability improvement SRR2 http://www-robotics.jpl.nasa.gov/ 5/33 Rocky 7 http://www-robotics.jpl.nasa.gov/ Robots with tracks Helios et Genryu http://www-robot.mes.titech.ac.jp
Design process of a climbing robot Purpose of the work Designing innovative robots with advanced wheeled mobility A general platform of wheeled vehicles : the Open WHEEL project Quasi-static climbing of obstacles Minimal number of actuators Proposed strategy 6/33 1. Design workspace definition Which type / architecture of mechanism? 2. Structural synthesis Which mechanism? Mobility analysis 3. Dimensional synthesis Which optimal dimensions? (not covered in this work)
Proposed design workspace Open WHEEL Open -> Open architecture for advanced mechanical concepts WHEEL -> Keeping the good efficiency of the wheel A generic framework for building a family of wheeled robots with canonical components and axle-based architecture Wireless connection A3 Rear CAN Bus W 32 S 32 Control S 31 Wheel W31 A2 I2 S 21 W21 7/33 Swing arm Double wishbone Innovative suspension W12 Camera S22 S 12 A1 Control Suspension mechanism Saw W22 Control I1 S 11 W11 nt Fro Z X Y Inter-axle mechanism Ia Serial Parallel Innovative mechanism mechanism mechanism
Structural design Key ideas Idea1 : a chair with 3 feet is always stable Idea2 : 4 wheels are good Exploring wheel Tricycle : dynamic stability problems Side-car : directional problems Conclusion 8/33 Design a robot with 3+1 wheels Dynamic stability on 4 wheels at high speed Always stable on 3 wheels when climbing Permanent static stability via adjustment of the centre of mass One 'exploring' wheel goes on top of the obstacle Support triangle V v N o
Structural design Mobility analysis The exploring wheel W11 should have an ascending movement in the plane Π 11 X translation for bringing the wheel towards the obstacle Z translation for lifting the wheel over the obstacle Y translation allowed during lifting but the track width of the vehicle should not change after landing Exploring wheel W22 Z Y 9/33 W12 S22 S12 A2 Re I1 ar S21 X A1 W11 S11 W21 Possible climbing trajectories
Example: Open WHEEL i3r Naming convention Naming the Inter-Axle Mechanisms Naming the Suspension Mechanisms Ex : irrr or i3r Ex : srpr Concepts of Open WHEEL i3r 10/33 Only one central R actuator allows to lift the four wheels according to mass repartition [IROS 2006] Capable to climb step obstacles via a specific climbing process All the work is on the inter-axle mechanism. No suspension. ar Re W21 W22 t n Fro W11 W12
Structural design Mobility analysis & suspension leg synthesis (a) (b) (c) (d) Suspension design : many mechanisms allow the required frontward-upward movement Current work in progress for extensive enumeration s4r : high DOF but low stiffness srpr : needs two DOF to position the wheel center spr : only one DOF if a angle allows ascending movement srr : frontward-upward movement is obtained in the south-east part of the circular trajectory a) s4r legleg (a) s4r b) srpr legleg (b) srpr R c) spr (c) sprleg leg -90 < <0 R P R 11/33 0 < <90 P R R d) srr (d) srrleg leg R R R R
Structural design A common architecture : 4sRR Gofor, JPL, 1992 12/33 NanoRover, JPL, 1994 ar e R OpenWheel 4sRR, 2005 W22 Z t Y W21 W12 W11 X n Fro Our goal : defining a climbing strategy
Designing a climbing sequence Design principles Gg = mc. Gc 4ml. Gl 4mw. Gw 13/33 Find the sequence of leg positions that allow stable climbing Vehicle can be on four wheels or only on three wheels The global centre of mass Gg is updated mc 4m l 4mw The support polygon is updated too The stability margin S is computed at every position S = min (Si ) from to i=1 i=3 or 4
Design tool : GeoGebra Algebraic & geometric tools -> 2,5D model Geometrical construction tools Computing the centers of masses Geometrical model Side view Analytical expressions Handles for interactive location adjustment A 2.5 D Model Adjustable design parameters Geometrical model Top view Construction protocol 14/33
A climbing sequence Structure 16 stages separated by single functional motions 6 phases from A to F Successive lifting of the four wheels 01 A 05 04 03 02 B 09 08 07 06 C 12 11 10 D 15 14 13 E 15/33 16 F
Stage 01 Detailed views 16/33 Side view Top view Real model A - Approach
Stage 02 17/33 B Wfl climbing
Stage 03 18/33 B Wfl climbing
Stage 04 19/33 B Wfl climbing
Stage 05 20/33 B Wfl climbing
Stage 06 21/33 C Wfr climbing
Stage 07 22/33 C Wfr climbing
Stage 08 23/33 C Wfr climbing
Stage 09 24/33 C Wfr climbing
Stage 10 25/33 D Wrr climbing
Stage 11 26/33 D Wrr climbing
Stage 12 27/33 D Wrr climbing
Stage 13 28/33 E Wrl climbing
Stage 14 29/33 E Wrl climbing
Stage 15 30/33 E Wrl climbing
Stage 16 31/33 F Conclusion
Conclusion Conclusions A general method for designing and modeling mobile robots with an optimal obstacle-climbing mode for a single step Design process in three stages: 1) Design workspace is the OpenWHEEL framework (axle-based architecture) 2) Structural synthesis showed the interest of the 4sRR kinematics for climbing 3) Dimensional synthesis (future work) 32/33 An original tool mixing algebra and geometry (Geogebra) proved its efficiency for preliminary design A 2.5D model and a reduced model were built A climbing sequence was created and demonstrated
Future work Four tracks for the future Explore all the feasible kinematics i3r 4sRR ar Re W21 W22 n Fro W11 t W12 Re ar W22 Z t Y W21 Innovative frame W12 W11 X n Fro Explore the other climbing sequences Example : After lifting Wfr, should we lift Wrr or Wrl? Design branching 33/33 Stage 14 Optimizing the climbing sequence: optimizing simultaneously structural and kinematical parameters on the 16 stages -> Tough! Solution: Bringing Gg forward? Minimizing the number of actuators: using one actuator for two revolute joints Stage 03