Research
Omata Lab studies grasping/manipulation and novel topics developed from them.
Our work started with dexterous manipulation with a multifingered hand.
Then we have been developing robot hands that can exert a very large fingertip force.
Recently, we are studying medical robotics in collaboration
with Tokyo Medical and Dentist University to solve problems in laparoscopic surgery.
Since surgeons cannot manipulate internal organs or tissue directly with their hands,
contribution of robot hand technologies is highly expected.
Novel surgical instruments that can be assembled in the abdominal cavity are proposed.
Legged robots should be able to grasp/manipulate an object.
We have developed a quadruped robot that can grasp/manipulate an object using its whole body.
The study on grasping/manipulation leads to the idea of grasping/manipulating a robot itself.
We have proposed self-reconfigurable parallel robots that can change their configurations by coupling and decoupling,
and also proposed parallel mechanism mobile robots that can form a three-dimensional structure.
Surgical instruments that can be assembled in the abdominal cavity for laparoscopic surgery
In laparoscopic surgery, surgeons perform surgery in the abdominal cavity
by using instruments inserted through trocars which are typically smaller than 12mm in diameter.
This surgery does not make a large incision, and therefore patients can make quick recovery.
However, instruments for laparoscopic surgery are slender because they need to be inserted through trocars.
Large traditional instruments for abdominal surgery cannot be used.
To overcome this problem, we have proposed assemblable instruments for laparoscopic surgery
that can be disassembled to pass through a trocar
and assembled inside the abdominal cavity to become large instruments.
Pursestring suturing as shown in Figure 1 is an application of this idea.
Fig. 1 Pursestring Suture
Fig. 2 PSI
In total gastrectomy, the esophagus and the duodenum (or the clone) are anastomosed typically
by using the anastomosing instrument as shown in Figure 1.
To fix an anvil in the esophagus, pursestring suturing of the cut end of the esophagus is necessary.
The traditional pursestring suture instrument shown in Figure 2 is T-shaped.
To insert this, an incision of typically 7 to 8 cm is necessary. Clearly this is not desirable.
Therefore, we have developed the assemblable pursestring suture instrument shown in Figure 3.
The introducer and the holder are inserted through two different torcars.
The holder holds the suture unit rigidly to become a T-shaped instrument.
Fig. 3 The process of connection(6.9MB Mpeg1)
To generalize this idea, various shaped instruments can be assembled in the abdominal cavity.
Much work has already been done to develop multi-degree-of-freedom forceps.
However they are still inadequate to handle or retract large internal organs, such as spleen, pancreas, and liver.
So we have proposed assemblable hands for laparoscopic surgery.
Figure 4 shows three-fingered five-degree-of-freedom assemblable hand.
As shown in Figure 4, we have experimentally verified that the hand can be assembled
in a closed box and can grasp large and oily objects like internal organs.
Fig. 4 Assembling and grasping in the closed space
There are two goals for this study.
One is to develop a simple hand with a few joints that has a single function
such as retracting or grasping, but still more useful than traditional retractors or graspers.
The other is to develop a skillful hand aiming at the replacement of surgeons' hands
in Hand Assisted Laparoscopic Surgery.
In HALS a human hand is inserted into the abdominal cavity through an incision of 7 to 8 cm,
which is invasive. The assemblable hand requires only small incisions.
Robot hands that can exert a strong fingetip force (CVT)
Fingers of a robot hand require speed when opening and closing and power when grasping an object.
However, it is difficult to improve the speed and power performances simultaneously
if the reduction ratio of a finger joint drive is constant.
A continuously variable transmission (CVT) is a solution to this problem
because low reduction ratios enable quick motions and high reduction ratios enable large force or torque.
The outputs of most existing CVTs are rotational motions of 360 degrees, such as V-belt CVT and toroidal CVT.
This is mainly because they were invented for automobiles.
However, robot joints do not necessarily rotate 360 degrees.
Focusing on this fact, we have developed a remarkably simple load-sensitive CVT
consisting of a five-bar linkage and a torsion coil spring.
The figure below illustrates the process of continuously varying transmission.
When the fingertip approaches the object, no external load is applied to the fingertip.
The length of the virtual input link Lvir maintains its maximum by the torsion coil spring in (a).
Therefore, the finger can move fastest with the lowest reduction ratio.
After the fingertip makes contact with the object,
the input torque needs to be increased to increase the fingertip force.
The length of the virtual input link Lvir becomes shorter and the reduction ratio R becomes higher
((b)-(d)) as the input torque is increased.
The developed finger can lift up a load of 2kg.
Quick and powerful motion(12.3MB)
We have succeeded in developing a 100g finger joint that can exert a 100N fingertip force
by considering the mechanical strength of its parts.
Quick and powerful motion of the new 100g-100N hand(17.1MB)
Dextrous manipulation with a multifingered hand
We studied dextrous manipulation with a four fingered hand. A human can
adjust his/her fingertip forces instantly when an external disturbance
is applied to a grasped object. A robot hand must be able to do this not
to drop a grasped object. We installed the grasp force optimization algorithm
developed in Technical University Munich. The four fingered hand in the
figure can adjust its fingertip forces very fast, every 4m second.
It can grasp an object firmly and can regrasp an object.
Another problem to be solved is to plan motions of fingers to achieve desired regrasping of an object.
This planning problem would be difficult to solve if it were formulated in "Configuration Space" because of its high dimensionality.
Our approach is to use primitive motions of regrasping which can be identified in the motions of fingers of human hands.
A sequence of the primitives to achieve desired regrasping can be easily planned.
Power grasps can grasp an object more firmly than fingertip grasps by using the whole of a hand.
The grasp force of a power grasp is in general indeterminate if friction is assumed in contact points.
No further analyses had been made on this point.
We analyzed the kinematics of power grasps with Prof. Joel W. Burdick, California Institute of Technology,
and showed an efficient algorithm to compute the bounded region of the indeterminate grasp force.
Whole quadruped manipulation
Legged robots are much more useful if they can perform manipulative tasks.
We have developed a quadruped robot Quadlator I which has a total of twelve joints.
Although its joint arrangement is typical for a quadruped robot,
its distinctive features are that it can stand on its knees and its two legs can serve as arms.
It can manipulate a relatively large and heavy object and can also perform tasks exerting a force on an external object
by using its body and arms.
The two legs standing on their knees support the body.
Therefore the motion of the body is constrained.
We proposed an appropriate contact mode for their distal links,
allowing the body to have a two degree-of-freedom motion
suitable for the whole body manipulation even if the floor or ground is not completely even.
Recently we have developed Quadlator II. It features:
(1) It is made self-contained, that is, all batteries, drivers, and computers are in its body.
(2) Its leg employs a parallelogram mechanism, which is suitable for walking.
Parallel mechanism mobile robots
We have proposed a crawler mobile robot whose upper body is a parallel
mechanism, 5R closed kinematic chain or 4R closed kinematic chain.
The robot has useful functionalities other than just locomotion:
Those are getting over a bump and carrying an object by making use of the parallel mechanism.
Furthermore, the robot can form three-dimensional structures
with other ones and go up to a high level to which the single one cannot.
Egetting over a vertical bump
Ecarrying an object
Egoing up to a high level
Self-reconfigurable planar parallel robots
We have proposed a novel self-reconfigurable parallel robot. In the figure,
the limbs (legs) of the proposed parallel robot are 2R (two-revolute-joint) open kinematic chains whose second joints are unactuated.
Reconfiguration is done by coupling and decoupling the platform and limbs, which enables the parallel robot to change workspaces.
The parallel robot can self-reconfigure without additional actuators by making use of its uncertainty singularity.
Home