IROS 2016 论文列表

This paper aims to develop novel surgical forceps integrated with a three-axis force sensor for robot-assisted minimally invasive surgery (RMIS). To detect accurate force sensing, a force sensing system is integrated to a gripper of a surgical instrument. At the gripper side, two possible locations are considered, and the sensing system is installed to the distal region of the gripper, which gives major advantages such as the gripper's minimization, palpation function, and multi-axis force sensing. Based on the capacitive-type sensing method, the sensor enables the direct measurement of the three-axis force applied to surgical gripper tip. The sensorized gripper is simply designed at a low cost, composed of only four mechanical parts, and, the forceps including two grippers are installed to an instrument that is able to conduct a grasping motion. Therefore, it is used to evaluate the performance of the force sensing system. The sensorized forceps are experimentally validated by using a commercial sensor in an experimental set-up.

There is minimally invasive method called Interventional Radiology (IR) that achieves an operation by using diagnostic imaging equipment such as Computed Tomography (CT). IR has advantages that it is a minimally invasive method and it is applied widely such as lung cancer treatment, liver cancer treatment, biopsy, and so on. However, it has a problem that doctors are exposed by X-ray due to a procedure in the vicinity of the CT gantry. To solve this problem, we have developed a robot so we calling “Zerobot”. Zerobot has six DOF and controlled remotely. We previously conducted animal experiment using a rabbit to evaluate Zerobot. As a result, percutaneous puncture was impossible without epidermis cutting. Therefore, we developed a new function called instantaneously puncture that can achieve high speed and short stroke puncture like a human operator. In this paper, overview of Zerobot is described first. Then, measurement of puncture speed of a human operator is described. Furthermore, design and system structure of instantaneously puncture system is described. Finally, we present results of animal experiment using the instantaneously puncture system.

In this paper, we present a design and development an autonomous tomato harvesting robot. We developed a harvesting robot with stereo camera which can measure depth in short range and direct sunlight and plucking gripper using the infinity rotational joint. We also evaluate the developed robot through harvesting in the tomato robot competition and the real farm. In the tomato robot competition, the robot harvested tometoes from tomato clusters and tomato trees, harvesting speed was about 80[s/fruit] and successful rate was about 60%. In the real farm we evaluated the robot with tomato trees in semi-outdoor environment to show the effectiveness and robustness under direct sunlight. According to the result of harvesting with real tomatoes, we improved the robot motion and finally harvesting speed was up to 23[s/fruit], however the gripper may grasp multiple fruits in case of very cluttered cluster and the calyx also may be broken when the stem angle is deep from the rotation axis. To avoid this situation, a grasp state estimation of the gripper and simultaneous recognition fruit and stem positions are next problems to improve the harvesting successful rate.

Tethered robots often experience entangling of their cables with obstacles in uncertain disaster environments. This paper proposes a spherical tether handling device that unfastens a robot's tether during surveys by releasing the tether and carrying it aside. By using a differential mechanism, the device drives shells and rollers that hold the tether. On flat surfaces, the device moves forward by driving the shells. When the device climbs over steps, the rollers are driven by the differential mechanism to pull the tether automatically. After prototyping the device, we confirm the surmountability of the proposed device against steps. The results show that the device can climb a height 90.9% of its diameter. We also demonstrate a scenario to handle the tether and untangle multiple tangles in an environment with several obstacles.

The tutor-tutee hand-in-hand teaching may be the most effective approach for a tutee to acquire new motor skills. Repetitive nature of such procedures in a group setting usually results in a high labour cost and time inefficiency. Potential solution can be utilizing robotic platforms playing the role of tutors for demonstrating and transferring the required skills. This requires an appropriate guidance scheme to integrate the tutor's motor functionalities into the robot's control architecture. For instance, for hand-in-hand supervision of the writing task, the tutor's corrections can be applied when necessary, while a very compliant motion can be achieved if no errors are detected. Inspired by this behavior, we develop a teaching interface using a dual-arm robotic platform. In our setup, one arm is connected to the tutees arm providing guidance through a variable stiffness control approach, and the other to the tutor to capture the motion and to feedback the tutees performance in a haptic manner. The reference stiffness for the tutors arm stiffness is estimated in real-time and replicated by the tutees robotic arm. Comparative experiments have been carried out on a dual-arm Baxter robot. The results imply that the human tutor is able to intuitively transfer writing skills to the tutee and also show superior learning performance over over some conventional teaching by demonstration techniques.

In this paper, we propose manipulation ability extension of the optical tweezers by developing microhands, which are to use as end-effectors of the laser beam. First, three different 3D micro-scale handles are designed, then manufactured by the two-photon polymerization method with nano-scale resolution of 100 nm. Second, printed microhands are manipulated by multi-spot laser beam which traps and manipulates numerous objects simultaneously. Third, where direct trapping of the target object is not possible due to target objects' features such as size, shape, material, index of refraction, etc., indirect manipulation of the target microobjects is achieved by using the microhands as an extension of optical tweezers. Finally, three different microhand designs are compared in terms of speed and success rate. Furthermore, suitability of different shapes of microhandles against common usage of spherical shape is discussed.

Targeted drug delivery by using magnetic nanoparticles (MNPs) is an efficient technique to deliver drug molecules towards specific tissues in a human body. The MNPs' 1D guidance system is a combined electromagnetic actuation (EMA) and monitoring system, which can provide an accurate control scheme with nanoparticle's localization for more precise targeting of the drug delivery. The localization of the MNPs is done by magnetic particle imaging (MPI) with low amplitude excitation field. In this paper, we have developed a novel coil topology for 1D MNPs feedback control, by alternate supply of different currents to coils set in time sequence, the coil set alternates functions between MPI and EMA. Motion of MNPs is controlled by a gradient of the magnetic field in EMA period, the distribution of MNPs is reconstructed in MPI period and provides feedback to the EMA. The guidance system will provide navigation and tracking interleaved for targeted drug delivery of MNPs in compact and efficient ways. The 1D MNPs guidance system has 2Hz of EMA and MPI hybrid frequency, allowing a position control of MNPs with 90nm diameter.

To enable robotic sensing for tasks with requirements on weight, size, and cost, we develop an ultra-tiny line laser range sensor based on the Time-of-Flight (TOF) principle. With delicate circuit design and optical attachments, we create a sensor as small as 35[mm] × 27[mm] × 30[mm] and as light as 20[g]. The line sensor samples 272 pixels (256 effective pixels) uniformly distributed within the measurement field of view customizable using different laser lenses. The optimal measurement range of the sensor is 0.05[m] ∼ 2[m]. Higher sampling rates can be achieved with a shorter range. The sensor can also extend its range to 3[m] with reduced accuracy. We model the overall errors of the sensor and formulate calibration methods, achieving repeatable accuracy and measurement bias both within 2[cm] with our tested ambient lighting conditions and measurement ranges. The sensor is applicable to range sensing tasks including humanoid hand-eye measurement, UAV safe landing, tiny robot range sensing, and object detection.

This paper presents a grasping force-feedback user interface (GFUI) which transfers the accurate kinesthetic force detected during robot-assisted minimally invasive surgery (RMIS) to a user, and allows the users to perform grasping motions with articulation and intuition. Using the GFUI, a developed robot hardware (S-surge) is controlled, and the force data measured from the robot are reflected to the GFUI. To evaluate the effect of grasping force-feedback control using the GFUI, a group of subjects (n=10) participated in a standard peg transfer tasks with and without the feedback control. To conduct the tasks, a developed robot hardware and a commercial haptic device were used which make the entire surgical environment including the GFUI as a master-slave controlled system. As a result, the performance of the GFUI was verified through the conducted tasks using the entire surgical system.

In this paper, a hand exoskeleton system for virtual reality is proposed. As a virtual reality interface for the hand, a wearable system should be able to measure finger joint angles and apply force feedback to the fingers with a simple and light design. In the proposed system, the finger joint angles are measured by a tendon-inspired cable mechanism. Also, another cable is used for force feedback to the finger. Using the measured finger joint angles and motor currents, the cable-driven actuation system applies desired force to the fingers. That is, when the desired force is zero, the motor position is controlled to track the changed cable length; when the desired force needs to be applied, the motor current is controlled to generate the desired force. For a smooth transition between two control strategies, linearly changing proportions of each control strategy is applied in the transition range. A prototype of the proposed system was manufactured, and the proposed control algorithms were verified by experiments.

Monitoring of biosignals plays important roles in health management in daily life. Especially, the sensing method of biosignals, which does not require special efforts such as restraining people or wearing the sensors to measure them, is really important to maintain monitoring activities of people. We call such method what a sensing way should be for monitoring of biosignals as casual sensing methods. Previously, we have developed highly sensitive and wide-measurement-range load sensors based on measuring frequency shift of quartz crystal resonator (QCR) as the sensing principal. We have integrated the load sensor into a chair, and measure the load when people just sit on it. Since the load sensor has wide measurement range of 105 order, multi-biosignal; heartbeat, respiration, and body motion under the weight-loaded environment can be measured from load information. Thus, people can casually monitor the biosignals by sitting on the chair. However, the previous sensing system required us to measure the frequency shift of one-Hz order in tens-MHz of the resonant frequency of a QCR. In this case, the sensing system requires a expensive frequency counter, and it is not suitable for daily use situation such as in-home sensing. In this paper, we presents the detection of multi-biosignal method using a newly developed load sensing system which utilizes the developed frequency counter unit. In order to measure the frequency shift of one-Hz order in tens-MHz of the resonant frequency, we use a differential method of signal for two QCRs with an electrical signal-mixing circuit. By using the method, we can reduce the required measurement range for frequency from tens-MHz to tens-kHz. The load sensing performances were evaluated, and the results showed that the sensitivity and withstand load were 2.9 [mN] and 300 [N], respectively. Finally, we demonstrated the measurement of multi-biosignal by using constructed system, and succeeded in detecting respiration, heartbeat, and bod... Show More

We present a novel and real-time method to detect object affordances from RGB-D images. Our method trains a deep Convolutional Neural Network (CNN) to learn deep features from the input data in an end-to-end manner. The CNN has an encoder-decoder architecture in order to obtain smooth label predictions. The input data are represented as multiple modalities to let the network learn the features more effectively. Our method sets a new benchmark on detecting object affordances, improving the accuracy by 20% in comparison with the state-of-the-art methods that use hand-designed geometric features. Furthermore, we apply our detection method on a full-size humanoid robot (WALK-MAN) to demonstrate that the robot is able to perform grasps after efficiently detecting the object affordances.

Autonomous vehicles are actively being developed from ADAS(Advanced Driver Assistance Systems) toward fully autonomous vehicles. Motion planning is one of the most important key technologies for fully autonomous vehicles, especially when they are operated in constrained narrow space such as parking lot. In this the motion planning is challenging because it requires many changes in forward and reverse directions and adjustments of position and orientation. In this paper, an efficient motion planning algorithm is proposed based on Rapidly-exploring Random Trees (RRT) by specifying desired orientation during the tree expansion. A tangential vector space for desired orientation is used to model nonholonomic constraints of a vehicle and geometric constraints of obstacles. The proposed algorithm has been tested on various situations and its results demonstrated much faster performance compared to a nonholonomic RRT algorithm.

In this paper, we propose a novel architecture to estimate external forces applied to a compliantly controlled balancing robot in simulations. We use similar dynamics equations used in the controller to find mismatches in the available sensory data and associate them to an unknown external force. Then by decomposing Jacobians, we search over the surface of all body links in the robot to find the force application point. By approximating link geometries with ellipsoids, we can derive analytic solutions to solve the search problem very fast in real time. The proposed approach is tested on a complex humanoid robot in simulations where it outperforms static estimators over fast dynamic motions. We foresee a lot of applications for this method especially in human-robot interactions where it can serve as a whole body virtual suit of tactile sensors. It can also be very useful in identifying the inertial properties of objects being manipulated or mounted on the robot like a backpack.

We propose a novel multi-rotor flying platform, ODAR (omni-directional aerial robot), which is fully-actuated (i.e., can assume arbitrary motion in SE(3) or generate arbitrary control wrench in se(3)) by six opportunistically distributed rotors, each driven by reversible ESC (electronic speed controller) for bi-directional thrust generation. Due to this omni-directional wrench generation, the ODAR system can realize such powerful behaviors impossible with conventional multi-rotor flying platforms as: 360° photo/video shooting while holding its position for VR scene generation; or resisting sideway gust while keeping its attitude and exerting downward pushing force larger than its own weight for aerial manipulation applications. This paper presents optimal mechanical design, modeling and control, hardware and software implementation, and experimental verification of the ability of the ODAR system to attain those behaviors impossible with the conventional drones as stated above.

The advantages of mechanical compliance have driven the development of devices using new smart materials. A new kind of magnetorheological piston based on a toroidal array of magnetorheological valves, has been previously tested to prove its feasibility. However, being an initial prototype its potential was still limited by its complex design, and low output force. This study presents the revisions done to the design with several improvements targeting key performance parameters. An improved annular piston design is also introduced as comparison with conventional devices. The toroidal and annular piston head prototypes are built and tested, and their force performance compared with the previous iteration. The experimental results show an overall performance improvement of the toroidal assembly. However, the force model used in the study still fails to accurately predict the magnetic flux at the gaps of the piston head. This deviation is later verify and corrected using a FEM analysis. The force performance of the new toroidal assembly is on par with the commonplace annular design. It also displays a more linear behaviour, at the expense of lower energy efficiency. Finally, it also shows potential for a greater degree of customisation to meet different system requirements.

Traditional climbing robots that use vacuum suckers have some technical problems, e.g., inability to climb coarse walls, frictional resistance and abrasion of suckers, and poor obstacle-surmounting ability. In this study, a new negative pressure adsorption mechanism is applied to the design of a climbing robot. This mechanism generates and maintains negative pressure and adsorption force by using the air's rotational inertia effect; therefore, the structure incorporating this mechanism is called the electrically activated rotational-flow adsorption unit. The most important characteristic of the adsorption unit is that it can function without being in contact with the wall, which fundamentally solves these technical problems associated with traditional climbing robots. In this study, we designed a square-shaped rotational-flow adsorption unit to improve the robot's load ability (18% increase in the adsorption force) and designed a soft skirt structure to improve the robot's obstacle-surmounting ability. Finally, we fabricated a prototype of the climbing robot and tested it on a actual wall (extremely coarse wall, wall containing a large groove). The test results show that our prototype robot can move stably on extremely coarse walls and can pass over large grooves easily.

The safety of working on walls has been improved by several adsorption mechanisms most of which sustain the absorption by pumps or other actuators. This research proposes an alternative method (a spring-suspended suction cup with a buckled plate spring) that can be attached for long durations without any vacuum pumping system. The negative pressure of the spring suspended suction cup is analyzed and the buckled plate spring is appraised by the energy efficiency and force over long adsorption times of the suction cup. The adhesion time of the suction cup is calculated and compared with that of the volume-fixed suction cup. Under high air inflow conditions, the spring-suspended suction cup adhered to the test surface for longer than the fixed-volume cup. The advantages of the spring-suspended suction cup were verified in experiments on a developed prototype.

This paper proposes a method to design low-cost robotic hands that require almost toolless assembly, and have special transmissions and joint mechanisms. The transmissions consist of active and passive gear trains manufactured by 3D printers to realize coordinated motions among joints and to transmit the driving force to underactuated joints. The joints are designed with pop-in structures to simplify the assembly of the designed robotic hands and its maintenance. The robotic hands even enabled us to exchange some components without necessitating a shutdown. By using these techniques, we manufactured two robotic hands with different passive gear trains. One of the hands had three fingers with stiff constrained gear trains and soft fingertips, and the other had five fingers with elastic gear trains. Experiments were conducted to validate the performance of the proposed robotic hands and to clarify the roles of the different gear trains.