Maplesoft is assisting Quanser Consulting to develop haptic technology with the aim of taking robot-assisted surgery to the next level. Haptics is a technology that allows human interfaces, such as joysticks, to provide force feedback cues to someone performing the procedure from a distance making him or her feel physically connected to the remote system. Robot-assisted surgery allows precision robotic tools to act like a surgeon’s arms, hands and fingers enabling surgeons to reach areas that would otherwise require large incisions.
Current haptic interfaces have limited or non-driven degrees of freedom (DOF). Quanser’s 5- and 6-DOF haptic interfaces (x, y, z, pitch, roll, yaw) featuring high stiffness, high response and minimal non-linearities meet the requirements for successfully integrating HMIs into real-life biomedical engineering systems (such as medical robotics).
Each DOF in every high-fidelity Quanser haptic device needs at least one servomotor set consisting of a precision motor and encoder, which are all connected with increasingly complexity in the form of a pantograph. As the DOF number increases, the complexity of the mechanisms providing the system inputs becomes so extreme that they are virtually impossible to model without resort to special software.
Quanser creates complex mathematical computer models to represent a robot’s motion, and uses the predicted behaviour to develop the controllers both for the robot motion and for the haptic feedback to the surgeon’s hands. In order to develop the controllers, Quanser modelled the behaviour of the mechanism using Maple mathematical modelling software.
The basic geometric relationships that describe the pantograph are entered and the software develops the systems of differential equations – many of which were several pages long – that model the kinematics and dynamics of the system. This is almost instantaneous and is completely free of errors typically inherent in manual mathematical manipulations.
Once this was done, the Quanser team was quickly able to test the model by solving the equations of motion and then developing the control strategies within the Maple environment. The team also used Maple’s optimised C code generation feature to export the model to Simulink as S-function blocks for final simulation and testing before building the controllers.