Page 78 -

Ch13-I044963.fm Page 61 Tuesday, August 1, 2006 12:49 PM
Page 61
Tuesday, August 1,2006
12:49 PM
Ch13-I044963.fm
61 61
Basic performance evaluation of the forceps manipulator; working range, force, torque, and speed, was conducted
(TABLE 1, Figure 6). The speed was measured under two conditions: one was conducted with the acceleration
time of 25[msec], the other was 1000[msec]. This was because the maximum speed of a stepper motor depended
on the acceleration time. We aimed to set the system control frequency to be 10[Hz] (100[msec/cycle]), and we
assumed that the acceleration time was at most quarter of cycle, 25[msec]. The speed of 1000[msec] acceleration
was measured as a case of enough long acceleration time. The working space was sector form whose radius was
340[mm] and whose vertex angel was 180[deg] in horizontal plane, and vertical depth was 360[mm] (Figure 6).

Bending forceps
We also measured the working range, positioning accuracy, backlash, maximum speed, and torque of bending
forceps (TABLE 2, Figure 6). Bending angle for "Bendingl" and "Bending2" was 304[deg] and 201 [deg] respec-
tively. The jaw for grasping opened up to 201 [deg]. Low positioning accuracy was caused because of not only
wire elongation and tension decrease, but also loose knot during the measurement. Output torque of "Bendingl",
"Jawl", and "Jaw2" were equivalent to the force; 1.9, 3.0, and 3.7[N] respectively. We found abrasion and break
of the wire, showing that friction occurred between the wire and path or pulley and that the transmission efficiency
was reduced by the friction.

In-vivo experiments
We conducted in-vivo experiments on a swine to evaluate the system in simulated clinical environment. Setting-up
time was less than 30 minutes and we thought it clinically feasible. The slave robot was controlled using a master
manipulator by a surgeon (Mitsuishi M, et al. (2003)). Each DOF had enough working range, however, we found
some problems. One problem was that; we inserted the surgical tool through trocar, an outer tube with air sealing
sleeve to avoid gas leakage. This air sealing fitted the shaft of forceps tightly, so that inserting motion was ob-
structed. We also had a problem about the direction of trocar. The initial direction of trocar was not necessarily
directed toward the target. In such a case, friction force between inner wall of trocar and outer surface of forceps
disturbed the motion of forceps. This issue will be solved by integrating trocar into the forceps manipulator.

DISCUSSIONS AND CONCLUSIONS
We developed a new compact robotic system as a slave robot in a master-slave system. Tt consisted of three mod-
ules; manipulator positioning arm, forceps manipulator, and robotized forceps with a two-DOF bending joint and
a grasper. Manipulator positioning arm realized intuitive easy setting up by the combination of rough and precise
positioning. Forceps manipulator realized four-DOF motion of the forceps around the incision hole with wide
working space (Figure 6). As for a RCM mechanism, "two liner actuator mechanism" realized mechanical

's$r-.

Figure 6: Working space; (left) forceps manipulator, (right) bending forceps