We study the average cost Linear Quadratic (LQ) control problem with unknown model parameters, also known as the adaptive control problem in the control community. We design an algorithm and prove that apart from logarithmic factors its regret up to time T is O( p T ). Unlike previous approaches that use a forced-exploration scheme, we construct a high-probability condence set around the model parameters and design an algorithm that plays optimistically with respect to this condence set. The construction of the condence set is based on the recent results from online least-squares estimation and leads to improved worst-case regret bound for the proposed algorithm. To the best of our knowledge this is the the rst time that a regret bound is derived for the LQ control problem.

/pdf/regret-bounds-for-the-adaptive-control-of-linear-quadratic-4r60v290d0.pdf

Regret Bounds for the Adaptive Control of Linear Quadratic Systems

In this paper, a spiral-type medical robot based on an endoscopic capsule was propelled in a fluidic and tubular environment using electromagnetic actuation. Both modeling and experimental methods have been employed to characterize the propulsion of the robotic capsule. The experiments were performed not only in a simulated environment (vinyl tube filled with silicone oil) but also in a real small intestine. The effects of the spiral parameters including lead, spiral height, the number of spirals, and cross section of the spirals on the propulsion efficiency of the robot are investigated. Based on the transmission efficiency from rotation to translation as well as the balancing of the microrobot in operation, it is demonstrated that the robot with two spirals could provide the best propulsion performance when its lead is slightly smaller than the perimeter of the capsule. As for the spiral height, it is better to use a larger one as long as the intestine's size allows. Based on the simulation and experimental results presented, this study quantifies the influence of the spiral structure on the capsule's propulsion. It provides a helpful reference for the design and optimization of the traction topology of the microrobot navigating inside the mucus-filled small intestine.

Modeling and Experimental Characterization of Propulsion of a Spiral-Type Microrobot for Medical Use in Gastrointestinal Tract

This paper presents a novel device that propagates through small pipes, named the twisted bundled tube locomotive device. The device consists of three elastic silicone rubber tubes that are twisted and bonded. When any tube is pressurized, it inflates, and the device deforms into a helix. If these tubes are sequentially pressurized, the device exhibits helical rotation, that is, it rotates about its body axis while retaining its helical shape. Thus, the device moves along the pipe axis. Since helical rotation is generated by bending motion alone, the device is highly airtight, and thus strongly resistant to the environment. Helical rotating motion does not scratch the inner wall of the pipe or block the flow in the pipe. The device can be made sufficiently small to travel through thin pipes, because of its simple structure. The device is very flexible and can pass through an elbow-pipe without requiring complicated control. Therefore, the device could be adopted in endoscopes used for pipe inspection. The propulsion principle and the theoretical velocity of the device are discussed and confirmed in prototypes. The manufacturing method for the prototype is also presented. The small prototype (approximate diameter 6 mm) travels at 45 mm/s in a 20-mm diameter pipe, and it passes through the elbow. Moreover, a prototype equipped with a camera can send images of the camera view as it moves the pipe.

A Twisted Bundled Tube Locomotive Device Proposed for In-Pipe Mobile Robot

This paper describes a state estimation model for a multi-segment continuum robot that utilizes the displacement of passive cables embedded along the robot's length to estimate its overall shape. As continuum robots are used in activities outside a laboratory environment, methods of measuring their shape configuration in real-time will be necessary to ensure robust closed-loop control. However, because these robots deform along their entire length and lack discrete joints at which primary displacements take place, conventional approaches to sensing joint displacement (e.g., encoders) are inappropriate. Furthermore, elasticity plays a key role in determining the resulting shape of the continuum robot, instead of the mechanics-independent kinematic configuration frequently seen in rigid-link robotics. In order to enable accurate estimates of a continuum robot's shape, the measured displacements of passive cables are utilized to detect the change in shape of the continuum robot. An optimization is used with a static model based on the principle of virtual power to map these cable displacements into the resulting continuum robot configuration. This state estimation model was implemented numerically in MATLAB and validated on an experimental test platform.

Multi-segment continuum robot shape estimation using passive cable displacement

Colorectal cancer is the third most common cancer in men and the second most common cancer in women worldwide. Diagnosing colorectal has been increasingly successful due to advances in technology. Flexible endoscopy is considered to be an effective method for early diagnosis and treatment of gastrointestinal cancer, making it a popular choice for screening programs. However, millions of people who may benefit from endoscopic colorectal cancer screening fail to have the procedure performed. Main reasons include psychological barriers due to the indignity of the procedure, fear of procedure related pain, bowel preparation discomfort, and potential need for sedation. Therefore, an urgent need for new technologies addressing these issues clearly exists. In this review, we discuss a set of advanced endoscopic technologies for colorectal cancer screening that are either already available or close to clinical trial. In particular, we focus on visual-inspection-only advanced flexible colonoscopes, interventional colonoscopes with alternative propulsion mechanisms, wireless capsule colonoscopy, and technologies for intraprocedural bowel cleansing. Many of these devices have the potential to reduce exam related patient discomfort, obviate the need for sedation, increase diagnostic yield, reduce learning curves, improve access to screening, and possibly avert the need for a bowel preparation.

/pdf/advanced-endoscopic-technologies-for-colorectal-cancer-o0cn1zgyox.pdf

Advanced endoscopic technologies for colorectal cancer screening

Purpose
Fibre optic colonoscopy is usually performed with manual introduction and advancement of the endoscope, but there is potential for a robot capable of locomoting autonomously from the rectum to the caecum. A prototype robot was designed and tested.

Development of a colon endoscope robot that adjusts its locomotion through the use of reinforcement learning

In this paper, the shape of the medical robot which can move in the colon is suggested for its gate control. In this system, the current shape information plays a sensing role in order to control the gate of robot in the colon. In order to find current shape of robot, we construct sensor network system which composed of several electronic compass units. This unit makes use of chip which includes pair of 3 axis accelerometer and 3 axis magnetometer. From this signals, orientation is evaluated after filtering noise. Then, based on the kinematic chain model, the shape of the flexible robot is calculated using orientation information. The resulting trajectory shows that this method cans percept shape of flexible robot well.

Non visual sensor based shape perception method for gait control of flexible colonoscopy robot

The general characteristics of the colon are proposed in the mechanical point of view. This is prerequisite information when we want to develop endoscopic robot which can move well in the human colon. Friction property which comprise static and dynamic friction coefficient is investigated. The elastic property of colon is also investigated: the radial tensile force is measured in the elastic range. The effect on the robot is revealed by using experimental method. The design parameters of the robot are then investigated: one is for the moving mechanism which has to be selected in view of mobility compared to the above knowledge. The other is for selection problem on the fin type and size in the inverse screw type. Several apparatus for friction property measurements were made and carefully made experiments. The resulting data were analyzed. Final results show that inverse screw type of robot has more mobility compared to the rotational inertia type.

Analysis of the colon by the biodynamic model and application to the colonoscope robot design

3D Visualization method of shape of flexible colonoscopy in simulated computing environment is proposed. Signals from sensor network are used to calculate the orientation of the body of each sensor. Position of each sensor is estimated from the orientation using assumption of forward kinematics in robotics. This data are then interpolated with curve fitting method to give natural impression of shape to surgeon who are watching monitor. The resulting simulated curve shows that motion of shape of colonoscopy can be simulated at it is.

Jaewoo Lee

Papers

Development of a colon endoscope robot that adjusts its locomotion through the use of reinforcement learning

Non visual sensor based shape perception method for gait control of flexible colonoscopy robot

Analysis of the colon by the biodynamic model and application to the colonoscope robot design

Shape visualization method of flexible colonoscopy using non visual sensor network for monitoring of operation