Marie Curie International Reintegration Grants (IRG)

pressdeadmancrossingAI and Robotics

Nov 14, 2013 (4 years and 7 months ago)


STREAM – PIRG03-GA-2008-231045 – Periodic Report

-- Scaled TeleRobotics for Enhanced Microsurgery --

Marie Curie International Reintegration Grants (IRG)
Grant Agreement No. PIRG03-GA-2008-231045

Publishable Summary:
Although teleoperation is around already quite some time (Geortz 1952
) it is still being
perceived as a very exotic and fascinating technological tool by the general public. Several
applications have shown how through inclusion of the ‘human-in-the-loop’ a robotic system
is able to execute far more complex tasks in uncertain environmental conditions than ‘stand-
alone’ robots can do up till now. Even more fascinating is perhaps their ability to overcome
physical barriers which humans encounter in daily life: barriers in space, barriers in time and
barriers in scale. Indeed, due to the physical separation between the slave robot working in
the remote environment and the master robot commanded by the human operator (Geortz
), operations can be conducted over a large distance (barrier in space) allowing access
to remote places (under water, in space, nuclear powerplants etc.). Operations of skilled
experts can be recorded, analysed and replayed later on (overcoming the barrier in time).
And also the barrier of scale can be relaxed. Motion and force commands at the master side
can be scaled up towards large motions and forces executed by the slave robot or scaled
down towards precise motions and gentle forces applied by the slave. While modest scale
factors could be realised by isomorphic master and slave robots, in order to achieve large
scale factors (both up and downscaling) dissimilar robot configurations/designs might be
The objective of STREAM, “Scaled TeleRobotics for Enhanced Microsurgery”, is to develop a
general framework that allows one to relax the ‘barrier of scale’. This framework is
demonstrated on a microsurgical application namely on retinal eye-surgery.

R.C.Geortz, “Fundamentals of General-Purpose Remote Manipulators”, Nucleonics, Vol.10, No.11, pp.36-42, 1952.
R.C. Geortz, “Electronically Controlled Manipulator”, Nucleonics, Vol.12, No.11, pp.46-47, 1954.

STREAM – PIRG03-GA-2008-231045 – Periodic Report
By choosing for a teleoperation approach we already have access to the
knowledge/intelligence of the expert surgeon. On top of this STREAM embeds in new ways
additional ‘intelligence’ into the robot controller. Clearly safe robot behaviour is a
prerequisite for robot surgery. The new framework allows one to deal in a deterministic
manner with the uncertainty introduced through unknown operator and environment
dynamics. This results in teleoperation controllers that behave sufficiently robust and
predictable. Several contributions were also made –under the denominator of Bounded
Environment Passivity - to improve controller performance. The newly developed Bounded
Environment Passivity approach allows one to reduce conservatism by including more
accurately the just amount of available operator/environment knowledge into the controller
design. Safe surgery execution further depends on hardware and familiarity with this
hardware. STREAM also pays special attention to these two aspects by developing dedicated
hardware, testing different possible operation modes and building up a realistic simulation
environment. The surgeon can exercise here; he/she can experiment with different
operation modes and tune the teleoperation controllers to his/her liking prior to going to the
real surgery.
In short STREAM’s objectives are summarized as
• developing a framework for robust bilateral control,
• developing a framework for robust variable-scale bilateral control and
• to develop a realistic training environment.
All these items will be demonstrated on an eye-surgical setup for which dedicated hardware
is being designed and built.
At present the framework for robust bilateral control is tested and implemented on a 1
degree of freedom (d.o.f) teleoperation system. The theoretic framework was extended
towards multiple d.o.f. systems. Also, in a number of publications, a novel method to
incorporate operator/environment information was described. In parallel a new 4-d.o.f.
slave robot for eye-surgery was built and interfaced. Current efforts go towards the design of
a force sensor to measure interaction forces with the retinal surface of the eye. Once
accomplished, experimental validation of the bilateral multi-d.o.f. controllers can take place.
An important remaining task will be to evaluate usability of the new system setup. Attention
will go to devising intuitive, potentially user-specific, operation schemes (predefined or
online adaptive scale settings) for different aspects of the surgical task. A-priori
experimentation in the simulation environment –under construction- could prove extremely
useful here.