Page 230 - Concise Encyclopedia of Robotics
P. 230

Odometry
                            in Fig. 1. The displacement changes at a rate that depends on the speed.
                            As long as you move forward, the displacement increases. If you go back-
                            ward, the displacement decreases. Displacement can be either positive or
                            negative with respect to the starting point.
                              In  two  dimensions, say  in  a  room  or  over  the  surface  of Earth,
                            odometry is done by keeping constant track of velocity, which has com-
                            ponents of both speed and direction.
                              Imagine boating in the open sea, starting from an island.You know the
                            latitude and longitude of the island; you can measure your speed and direc-
                            tion constantly.You have a computer keep track of your speed and direction
                            from moment to moment. Then, after any length of time, the computer
                            can figure out where you are, based on past movements. It does this by
                            integrating both components of velocity (speed and direction) simulta-
                            neously over time. Sailors know this as ded reckoning (short for deductive
                            reckoning) of position.
                              A robot can use ded reckoning by having a microcomputer integrate its
                            forward speed and its compass direction independently.This is called double
                            integration. It is a rather sophisticated form of calculus, but a microcom-
                            puter can be programmed to do it easily. Figure 2 shows two-dimensional
                            odometry based on speed and compass direction. The velocity vectors
                                  V              North
                                    4
                                                               Finish
                                           V
                                             3
                                                           V
                                                            2
                            West                                      East

                                              V
                                               1
                                                                Path of
                                                                robot


                                       Start
                                                 South
                            Odometry—Fig. 2




                                                   
   225   226   227   228   229   230   231   232   233   234   235