Solid Media Transmission
An Innovative and Versatile Transmission
Ideal for use in the widest range of manipulators
From the high and rapidly changing magnetic fields (in MRI-guided robots)
to the ultra-high pressure Undersea
or the zero pressure Outer Space and beyond
From human-size to milli/micro/nano manipulators
What Is the Solid Media Transition and the motivation behind its invention.
After our earlier work in developing MR compatible interventional robots (esp the one for breast and the other for abdominal/spinal interventions) we formed a vision about the next generation of MR compatible robots from our Lab, Within this context, the Solid Media Transmission (SMT) was conceived at the early stages of constructing a robot for intracardiac interventions as part of the MIROS CPS (an NSF funded project). In response to the criteria listed below, the fluid medium was substituted with spheres and then spacers were added in-between them for improving efficiency.
Our Criteria for the Next Generation MR Compatible Manipulators:
Do Not Use Ultrasonic or Hydraulic or Pneumatic motors: Use Conventional shielded, filtered and/or secured Electromagnetic Motors (EM) positioned outside the gantry
Connect the EM's to the robot To Degrees-of-Freedom (DoF) with some form of tranmssion that is easily flexible and adjustable like the fluidic transmissions (hydraulic or pneumatic) but do not use fluids!
In general, make them such that no special Preparation or changes are needed the MR scanner Room: Place the robot control electronics inside the MR room (shielded) and ecven avoid using the filter plate (if such exists!)
(a) & (c) an SMT with spheres and (b) & (d) an SMT with spheres (white) interleaved with bushings (black)
(a) A one DoF SMT end effector for testing (note how the tubimng is bend), (b) its shielded control box for operation in the MR scanner suite and (c) the system.
A four degrees-of-freedom (4 DoF) SMT- powered manipulator for MR-guided neurologic or abdominal area interventions .
Features of SMT
SMT technology can reduce production cost and complexity and avoid other disadvantages of alternative methods (ultrasonic or fluidic actuators). SMT has the key properties of fluidic actuators (remote actuation and flexible routing), while exhibiting certain unique key features:
Fluidless: This key feature has far reaching consequences in the design of SMT-actuated devices in regard to compactness and mechanical simplicity:
(1) it does not require an actuation piston that is at least as long as the stroke, and
(2) actuation can be transmitted from the SMT media to the actuated component, or vice versa) at any location via slots in the channels or tubing.
Flexibility in design and production: Secondary to its fluidless feature, SMT offers new opportunities for alternative and simpler designs, and production. For example, straight and curved SMT channels can be embedded into the frame of a manipulator, which can be produced by CNC milling or 3D printing two layers and then bolting them together (See designs, videos, manuscripts and patents).
Environmentally versatile: Without the need for hydraulic fluids, SMT is well-suited for any application where leaks and hydraulic contamination are undesirable. In addition, it can be made less prone to environmental conditions. This opens opportunities in special industries such as the aerospace and oil/gas that require systems to function in near-vacuum, at cryospheric temperatures, or in high-pressure subsea conditions.
MK2: A 1-DoF SMT Bot - our early workhorse
The MK2 is a linear stage, a 1 DoF SMT-actuated manipulator used in experimental studies to systematically investigate and characterize the operation of the novel SMT.
The designs exemplify some fo the important features of SMT: (1) the conduits can simply be cut-out on two-layer frames and (2) actuated can be transmitted via slots - no concern of leaking!
(1) Upper layer
(2) Bottom layer
(3) Inner channel
(4) Screw hole
(5, 6) Channel ends
(7, 8) Tubing fitting and adapter
(9) Carriage
(10) Slot
(11) Piston
(12) Optical encoder strip
(13) and (14) Stop switches.
Representative SMT Characterization Studies
Example experimental studies at the MRLab included:
System Identification and tuning the PID
Effect of Sphere Material (friction) and Conduit Sheathing (expansion)
Tubing Material: Nylon, PTFE, and braided PTFE
Sheathing: Acrylic sheath, and Braided coat
Effect of Conduit Length
Lengths from 1 to 4m
Spheres of PTFE/6.35mm & 20 mm long Spacers
Response to Sinusoid Frequency
Frequencies 0.2 Hz to 1 Hz; Amplitude +/- 5 mm
Spheres of PTFE/6.35mm & 20 mm long Spacers
The MK2 is carried by a robotic arm that, as moves, chnahes the deployment of the SMT conduit of MK2 for studying the eeffects of different paths
MK3: A 4 DoF Parallel SMT Bot
This robot is a wonderful manifestation of how the SMT can be realized fo a complex 4 DoF parallel kinematic architecture entailing a plurality of innovations on how to engage and link and engage 4 bi-directional SMT lines to 4 independent DoF: two rotational and two linear.
Video of 3D Design (Left): the primary components of the MK3 SMT-Bot with a unique way of implementing the parallel kinematic structure: each plane contains a rotational DoF (actuated by one bi-directional SMT) and carries a linear one onto which is attached the interventional tool via passive bearing. What is exciting is that the linear DoF is also actuated by a highly innovative rotational DoF! Photos (middle and right): the MK3 SMT-Bot carrying a biopsy needle.
SMT-MK3_20170511.aviMovie Above: Laboratory testing of the 4-DoF SMT for inserting a needle into a tube (target) using visual servoing (cameras feed is in the lower two windows). The operator inserts the needle manually through the bushings of the end effector.
Photo in the Left: The 4DoF MK3 Bot (distant background) connected to 4-m long SMT Lines that are connected to the EM motors (foreground) deployed at the Lab for testing. Note the SMT-specific pistons in the foreground specially developed for bidirectional of a DoF via two SMT lines (alike in conventional fluidic actuators)
Some SMT Modeling with Mathematica
Some wonderful simulations by our colleagues Haoran Zhao and Aaron T. Becker
Compression Ratio of Spheres in a Curved Tube
Transmitting Force through a Tube Filled with Spheres and Spacers
SMT-Bots, Electromagnets and ... MRI, how do they fit?
SMT is composed of media (spheres, spacers & conduits) made of non-magnetic/ferrous material; as a consequence its presence and media actuation does affect the generation of MR images (no noise or signal artifacts, non-linearities or inhomogeneities).
We did use conventional EM (made by Maxon; seen in the photos above) placed outside the 5 Gauss line and eventually were proven quite MR safe & compatible:
no force was exerted on the motors sufficient to dislocation (motors + SMT-specific pistons manufactured of aluminum).
the power supplies and controllers were placed in an EM shielded box (see figure below) that sufficient to preserve the SNR when the motors were ON and operating in open or closed loop control. The closed loop controller was running on the Unit.
The Control Unit was inside the MR scanner room and was powered by a wall power outlet.
The User Server (that runs the .MR-based servoing or path planning) it connects to the Control Unit via an optical cable and can reside inside (beyond 5G) or outside the scanner room. During robot operation, from the Unit receives a stream of optical encoder signals, commands sent to the controllers from closed/open loop. It sends to the Unit, commands to actuate the closed/open loop
Our conclusion was that, with appropriate measures, conventional EM motors can eb used tinside the MR scanner room, and controlled to actuate an SMMT-powered robotic manipulator inside the gantry of the MRI scanner
(a) MRI zoomed into the marker. (b) - (d) MRI depicting marker at different positions as manipulator is actuated. (b) the dashed box is approximate SMT manipulator frame. (b) - (d) the vertical arrow points to the marker.
The well-shielded SMT Control Unit resides inside the MR scanner room, houses power supplies, controllers, optical encoder boards, safety switches, emergency switches etc. Bi-directional communication with the robot server (that runs MR-based planning and servoing & can be inside or outside the MR scanner room) is done via an optical cable
Patents
SMT mechanism
and
art of MRI guided robotic interventions
US9539058.pdfUS10136955.pdfUS9855103.pdfNSF STTR Funding
Announcement of the NSF STTR Award that supported the further development of this technology
UH Startup Earns Commercialization Grant for New Technology
The People
A great team of amazing Engineers and Scientist worked closely under the support of the support of an STTR NSF and brought this project to fruition:
Dr. Michael J. Heffernan, PhD; a great Scientist with superb Management abilities and experience
Prof. Aaron Becker, PhD; a true Engineer who run the Engineering
Dr. Haoran Zao, PhD, a ECE who did modeling, simulations, construction and testing
Dr. Xin Liu, PhD; a CS who both running extensive experimental studies and developed software for MR servoing and path planning; a software engineer who learned MRI, robotics and their combination!
Mr Korpu who contributed major effort in bulding, assembl;ing and texting
Ms Yolanda who volunteered a her expertise and time in 3D designing!
The different aspects of this work were supported in part by the National Science Foundation Grants CNS-0932272, IIP-1622946 and CNS-1646566. All opinions and conclusions or recommendations expressed in this web page k reflect the views of authors not our sponsors.