In the 1980s, when the author worked for Seiko Epson Corporation as a wristwatch production engineer, consumer needs had become so diversified that wristwatches had to be assembled on the same automated assembly line in small lots of about 10,000 pieces per month. Most of the robots available in those days were for processing purposes such as spot welding and were not applicable in practical terms to automated assembly lines for wristwatches in precision, speed, ease of use or cost. The prototype SCARA robot developed by the SCARA Study Group led by Dr. Hiroshi Makino, a professor at the Department of Precision Engineering at Yamanashi University, was found to be the most suitable for automated watch assembly lines. We reviewed assembly work procedures and succeeded in limiting the number of simultaneous control axes to four at a maximum and in cutting assembly costs to 60% of those of conventional processing robots. As the term “selective compliance” suggests, SCARA robots possess all of the functions necessary for stable assembly and for precision and speed. Development team members, including the author, made the most use of previous experience in developing dedicated automated assembly machines for in-house use and succeeded in developing practical SCARA robots by creating a robot language based on workers’ voice. In applications of SCARA robots, this paper introduces just two of many possible examples. One is for oiling work on manual assembly lines and the other is the TAF-M mixed-models wristwatch assembly line. In the oiling work application example, SCARA robots used for infinitesimal oiling work on a manual assembly line for small lots of luxury wristwatches have been found to be very cost-effective and useful for training operators and/or programmers for robots. The TAF-M application example represents the assembly line built based on the original Seiko-Epson purpose for introducing SCARA robot development, which consists of both robots and of 52 newly developed “assembly robot cells” where most assembly work should be done. An assembly robot cell includes a SCARA robot, a main conveyor, a multiple-parts feeder, an automatic hand changer, an assembly detection unit, etc. At present, one such assembly line automatically assembles more than 100 models of wristwatches. Use of such automated assembly lines has reduced the human workforce by about 40 workers, cut costs by over 60%, and shortened delivery time by about 50%.

Almost four decades ago, Denso developed a pioneering high-speed fully automated line for meter gauges. In the years since, it has also developed several key production-line devices as production lines underwent dramatic changes. Denso’s original production line even now serves as a historically significant reference in the development of new production systems. We detail its development both for its historical value and as a guide to those now developing new process designs.

This paper presents a simple, natural and effective framework of nonlinear systems control and its application to aerial robots. First, we present a framework of Takagi-Sugeno fuzzy model-based control and also discuss its feature. Next, a number of design problems for the control framework are formulated as numerically feasibility problems of representing in terms of linear matrix inequalities. Finally, we provide two applications of the control framework to aerial robots. The control results of aerial robots show the utility of the control framework.

Robotics researchers appear to have shifted their focus since the Great East Japan Earthquake. Despite a large-scale national R&D project on Advanced Robots for Hazardous Environments for 8 years from 1983, the technology has not been put to use effectively following the Great Hanshin Earthquake and researchers are starting to look more closely at applications enabling robots to replace human beings, e.g., by rescuing victims of earthquake and flood disasters. A good example is Quince and the remote-controlled heavy equipment used to handle the pyroclastic flow at Unzen volcano in 1991 and then mobilized at the Fukushima Daiichi nuclear power plant following the Great East Japan Earthquake and tsunami in 2011. Robotics researchers of the past only used to describe their dream proposals when asked what robots could accomplish. Since the 2011 disaster event, however, it has become clear that the issues we should be looking at are those of developing robots that are practical and useful.

If we are asked what role the real-world competition Tsukuba Challenge plays in this context, would first recommend that those who hope to take part visit the actual site and see for themselves what they must do to field a winning robot. The Tsukuba Challenge site includes public thoroughfares traversed by pedestrians and cyclists. Although there are no cars or motorcycles on these paths, almost anything can happen. From hot summer until cooler fall when the official run is held daily temperature, rain, wind and typhoons, and trees – all of which must be detected for navigation – undergo many changes. These changes require that robots navigate accurately in this real-world environment. In general, robotics research papers are accepted for publication if robots can navigate as planned in a restricted and fixed mobile environment even if they can do so only once.

In the Tsukuba Challenge, however, many things can go wrong and robot maneuvers can become unstable or even nonmobile, regardless of successful navigation in laboratory settings. There is no space here to discuss all possible factors, but the established navigation method by one paper is not always successful and responsible in the actual Tsukuba Challenge environment. Robots historically came to be as devices operated by human beings, but those taking part in the Challenge must be controlled by computers instead of human operators. This means that it must be confirmed that robots can operate as required in experiments meeting various conditions. Teams may conduct ten or so trial runs a year on the actual route so that their systems can be adequately adjusted and modified. The total process that competing in the Tsukuba Challenge requires also has an educational effect in grooming new talent in robotics.

The Challenge differs from the DARPA Grand Challenge and Urban Challenge, held from 2004 to 2007, in the size – robots must not be too big and must not appear threatening – and the need to take the presence of human beings into consideration. Although the task from 2007 to 2012 had been only to navigate a preset route and reach a final goal, still the percentage of successful runs has declined. From 2013 on, another task has been added – that of locating specific persons (search targets) within given areas – to encourage the use of advanced technology in realizing useful robots. Those interested may see the results on the Tsukuba Challenge website as follow: http://www.tsukubachallenge.jp/tc2013.

The selection of articles for this special issue emphasized the following criteria:

1) For robots successfully completing required tasks, describing and discussing the superiority of the control technology and results.

2) For unsuccessful robots, clearly analyzing how the actual run differed from researchers’ expectations and pinpointing the underlying causes of failure.

The submitted papers...

[View PDF for the rest of the abstract]

In Real World Robot Challenge 2013, a mission was added that had robots search for a person wearing clothes featuring unique colors. We focus on the layout of such clothes with the aim of detecting persons wearing them by applying color extraction. Color extraction is improved by preprocessing of a clipping image from the robot’s vision and possibly extracting colors worn by target persons stably in natural light. Persons are detected by simply evaluating the layout of target colors. Our robots were equipped with person detection for this challenge and have detected all targeted persons. This paper describes considerations about person detection performance based on pre- and postchallenge results.

For the Real World Robot Challenge (RWRC) 2013, a new task was established: every robot was required to search for designated persons. In this paper, therefore, we consider the difficulty of the task and construct a navigation strategy to achieve the task. To navigate a robot on the basis of the strategy, long distance navigation is necessary. We have developed a unique navigation method based on magnetic and geometric landmarks on environmental structures in various locations. This method allows a robot to robustly localize by evaluating the reliability of magnetic and geometric landmarks. By using this method, a robot can navigate stably, even if there are no existing landmarks to serve as objects. We achieved autonomous navigation over long distances and successfully searched out designated persons as the challenge of the RWRC2013. This paper presents our navigation method and discusses long distance navigation using the method.

This paper describes a navigation method for autonomous mobile robots and the knowledge obtained through trial runs conducted during the Tsukuba Challenge 2013 whose main tasks were autonomous navigation by robots to a goal and searching for target persons in several urban areas. Accurate maps are an important tool in localization on complex courses. We constructed occupancy grid map making method using laser scanners, gyro-assisted odometry and a DGPS. In trial runs, robots detect target parsons two ways – one involving color detection in images and the other involving laser scanner intensity data. The major problem with these methods is misdetection. To minimize this, we mask areas in which target persons should not exist on occupancy grid maps. Target candidates detected in masked areas are rejected, which indicates the possibility of using accurate occupancy grid maps as a user-friendly graphical interface. This paper focuses on the localization method, the target detection method and autonomous navigation knowledge in common space through the challenge.

This paper describes a robust navigation method for real-world environments. The method uses a 3-axis magnetic sensor and a laser range scanner. The magnetic field that occurs in the environment is used as key landmarks in the proposed navigation method, and physical landmarks scanned by the laser range scanner are taken into account in compensating for the mobile robot’s lateral error. An evaluation experiment was conducted during the final run of the Real World Robot Challenge (RWRC) 2013, and the result showed that the mobile robot equipped with the proposed method robustly navigated a 1.6 km course.

This paper describes navigation systems for autonomous mobile robots taking part in the real-world Tsukuba Challenge 2013 robot competition. Tsukuba Challenge 2013 enables any information on the route to be collected beforehand and used on the day of the challenge. At the same time, however, autonomous mobile robots should function appropriately in daily human life even in areas where they have never been before. System thus need not capture pre-driving details. We analyzed traverses in complex urban areas without prior environmental information using light detection and ranging (LIDAR). We also determined robot status, such as its position and orientation using the gauss maps derived from LIDAR without gyro sensors. Dead reckoning was combined with wheel odometry and orientation from above. We corrected 2D robot poses by matching electronics maps from the Web. Because drift inevitably causes errors, slippage and failure, etc., our robot also traced waypoints derived beforehand from the same electronics map, so localization is consistent even if we do not drive through an area ahead of time. Trajectory candidates are generated along global planning routes based on these waypoints and an optimal trajectory is selected. Tsukuba Challenge 2013 required that robots find specified human targets indicated by features released on the Web. To find the target correctly without driving in Tsukuba beforehand, we searched for point cloud clusters similar to specified human targets based on predefined features. These point clouds were then projected on the camera image at the time, and we extracted points of interest such as SURF to apply fast appearance-based mapping (FAB-MAP). This enabled us to find specified targets highly accurately. To demonstrate the feasibility of our system, experiments were conducted over our university route and over that in the Tsukuba Challenge.

As in the Tsukuba Challenge, any robot that autonomously moves around outdoors must be capable of accurate self-localization. Among many existing methods for robot self-localization, the most widely used is for the robot to estimate its position by comparing it with prior map data actually acquired using its sensor while it moves around. Although we use such a self-localization method in this study, this paper proposes a new method to improve accuracy in robot self-localization. In environments with few detected objects, a lack of acquired data very likely will lead to a failure in map matching and to erroneous robot self-localization. Therefore, a method for robot self-localization that uses three-dimensional environment maps and gyro-odometry depending on the situation is proposed. Moreover, the effectiveness of the proposed method is confirmed by using data from the 2013 Tsukuba Challenge course.

This paper describes a localization method for mobile robots. The proposed method is based on a three dimensional space observation model that provides stochastic information for robot location assumption. We previously developed a localization method based on the two dimensional space observation model to be used in an environment with many unknown obstacles such an pedestrians. The previous method used a two-dimensional laser scanner and particle filtering to realize robust effective localization of the mobile robot in such environments. It sometimes failed to estimate the robots’ location due to a lack of information about its environment, so we extended the previous method based on a three-dimensional space observation model that is expected to use richer information about its environment. Experimental results showed that our proposed method localizes robots successfully in environments where the previous method could not localize robots accurately.

This paper describes the development of a mobile robot system and an outdoor navigation method based on global navigation satellite system (GNSS) in an autonomous mobile robot navigation challenge, called the Tsukuba Challenge, held in Tsukuba, Japan, in 2011 and 2012. The Tsukuba Challenge promotes practical technologies for autonomous mobile robots working in ordinary pedestrian environments. Many teams taking part in the Tsukuba Challenge used laser scanners to determine robot positions. GNSS was not used in localization because its positioning has multipath errors and problems in availability. We propose a technique for realizing multipath mitigation that uses an omnidirectional IR camera to exclude “invisible” satellites, i.e., those entirely obstructed by a building and whose direct waves therefore are not received. We applied GPS / dead reckoning (DR) integrated based on observation data from visible satellites determined by the IR camera. Positioning was evaluated during Tsukuba Challenge 2011 and 2012. Our robot ran the 1.4 km course autonomously and evaluation results confirmed the effectiveness of our proposed technique and the feasibility of its highly accurate positioning.

A design policy for autonomous mobile robots favors widely accepted using as many sensors and as much powerful recognition hardware as possible to realize reliable robot operation. If we plan to use developed technology for commercial products, a separate design policy favors a minimum number of sensors and recognition hardware, i.e., the number enough for reliable operation. We named the robot designed under the latter design policy the Minimal Autonomous Mover (MAM) and built a MAM to participate in the Tsukuba Challenge, a competition for among autonomous mobile robots. In this competition, our robot reached the goal and completed the mission as reported in the sections that follow.

This paper presents a path planning method by path energy minimizing that enables mobile robots to move smoothly in the real world with optimizing path shape for shortest distance or minimum curvature. It also enables robots to travel safely toward a destination because pedestrian motion prediction is embedded in path planning. This path planning method is based on problems experienced in a robot competition called Tsukuba Challenge. The problems involved nonsmooth motion arising from finite path patterns in A* algorithm, stuck motion arising from frequently path switching, and near misses arising from nonpredictive planning. Our path planning method minimizes path-shape energy defined as the connection between path points. Minimizing energy provides smooth paths and avoids path switching. We propose a path planning method with prediction of dynamic obstacle motion embedded to avoid near misses. Experimental results showed improvements in solving these problems.

We have been developing a position measurement system for navigation of automated guided vehicles (AGVs) called SPARS. In this system, the AGV’s ultrasonic position measurement module communicates via ultrasonic waves with ultrasonic transponders that serve as land markers on a path to measure its relative position during travel. In previous studies, we conducted experiments and introduced improvements using the relative position between the AGV and land marker estimated from position information from a single land marker. It was found, however, that the ultrasonic communication S/N ratio decreases, lowering position accuracy, when the land marker distance and its direction angle are great. To solve this problem and improve accuracy, we examine position measurement based on distance information from two land markers.

The purpose of this research is develop assistive robots and apparatuses. There is a pressing need to develop new systems that assist and act for car driving and wheelchairs for the elderly as the population ages. In developing systems, it is thought to be important to examine behaviors spatial recognition. Experiments have therefore been performed to examine human spatial perceptions, especially left- and right-side visual recognition, while cars being driven using near-infrared spectroscopy (NIRS). Previous research found significant differences in the dorsolateral prefrontal cortex in the left cranial hemisphere during virtual driving and actual driving tasks. This paper discusses the measurement of brain activity during car driving. A detailed analysis was performed by segmentalizing brain activity during driving based on the motion of subjects, and we report on the relationship between brain activity and movement perception during driving.^*

^* This paper has been reviewed and accepted for Special Issue on Assistive Technology Based on ICT/IRT for Aged Society, Journal of Robotics and Mechatronics, Vol.25, No.6, 2013.

A wheeled vehicle with crawlers between the front and back wheels has been developed in order to move swiftly over both soft ground and pavement. Vehicle trafficability on dry sand was evaluated in experiments, which clarified that its trafficability was greater than for wheels on dry sand alone.

In this article, a cruising autonomous underwater vehicle (AUV) r2D4 and examples of its achievements to date are introduced. With the objective of realizing an intelligent multirole platform for deep-sea surveys, the r2D4 development program was started in 2001. Launched in 2003, the r2D4 had achieved several practical missions by the time of its last dive in 2010. Enhanced vehicle autonomy was realized by applying new technologies in navigation, control, positioning, and sensing, which enabled the capabilities of tackling more challenging undersea missions. The r2D4 is a multirole platform that is easily able to be modified to meet the applications of diverse purposes. Since its first dive in September 2003, the r2D4 has successfully completed a total of 67 dives in bodies of water worldwide.

This force masking humanoid robot system had been developed for a psychological experiment that confirms the rhythmic modalities required by humans when they play rope turning with a robot. Although a person feels many types of rhythms in physical rhythmic cooperation with a robot, what rhythms are required for the cooperation had not been investigated yet. This system is an effective tool to investigate the effectiveness of the force rhythm in physical cooperations.

The objective of this study was to develop a humanoid robot for research about human swimming. In order to replace a swimmer as a subject, a humanoid robot “SWUMANOID” was developed imitating the details of a human, such as appearance and body properties. To reproduce the swimming motion based on that of an actual swimmer, a joint imitating the human’s scapular retraction was installed and the methodology to realize the swimming motion with that joint was established. To validate the performance of the robot, two experiments were conducted. The crawl stroke motion was examined in the fixed state and the swimming performance was validated by a free swimming experiment.

To replace manual labor, we developed an automated asparagus harvesting robot consisting of a proposed end effector, a robotic arm, a laser range sensor and a mobile cart for navigating asparagus furrows. The major feature of our robot are that it harvests asparagus speedily and works at whatever hours it is needed.

The passive assist arm (PAS-Arm) that we developed interacts with the operator by producing specified but arbitrary two-dimensional virtual guiding surfaces that constrain and guide the operator’s movement.