This paper introduces a sensing with YOLO and control technique on a humanoid climbing robot. Humanoid robots are expected to assist a variety of our tasks. We have focused on free climbing performed by a humanoid robot. In practice, the proposed controller solves the non-linear optimization problem to derive the best posture for climbing and perform visual servo control on an arm or leg. Moreover, this paper presents the robot's effectiveness with additional waist motion.

In order to realize high-level force control, detailed force information is indispensable. Therefore, we focused on photoelasticity, which enables us to obtain stress distribution at the pixel level of a polarization camera. In this study, we developed a robot hand with the fingers and palm mostly made of photoelastic material in order to obtain the stress distribution of the entire material with high spatial resolution. We used a polarization camera that can capture polarization images in four directions simultaneously, and performed image-based stress analysis. The experimental results showed that force distribution can be measured and the measured resultant forces were in good correspondence with those measured by a conventional force sensor.

In recent years, the manufacturing industry has been promoting the introduction of robotic cell production systems suitable for variable production, and in such systems, robots need to handle parts in various shapes and assemble them with high accuracy. In particular, when grasping three-dimensional (3-D) shaped parts, it is desirable to grasp the parts accurately without wobbling. In order to absorb the uncertainty of the initial pose of a target part and plan a robust grasping strategy by which the part can be grasped in a desired pose, it is necessary to evaluate the permissible initial pose error region (PIPER) of the part under an assumed grasping strategy. In this paper, we present a method to derive the PIPER of a 3-D shaped part for grasping with a parallel stick fingered hand based on the dynamic analysis of pushing operation considering wobbling of the part.

We are conducting research and development of mROS 2, that is an agent-less and lightweight runtime environment for ROS 2 node onto embedded devices. In this paper, we make the communication protocol stack of mROS 2 compliant with POSIX (Portable Operating System Interface). Since POSIX is a unified interface standard for operating systems, this work enables mROS 2 to operate easily onto Ubuntu OS, that is the default platform for ROS 2. We implemented the proposed method on Ubuntu 20.04 running on Raspberry Pi 4B. Experimental result showed that our research outcome could improve communication performance than the native ROS 2 environment.

In recent years, automation of construction and civil engineering industry sites is highly demanded. In particular, ``excavation and loading'' is an essential task that is achieved with a collaborative work between backhoes and dump trucks. For the automation of the work, backhoes need information of soil condition inside the truck bed such as volume, shape and position of the soil loaded in the bed. For example, overloaded soil causes collapse from the truck bed while moving. That is why these kinds of information is necessary for backhoes to achieve the loading task efficiently. In this paper, we propose a method to measure soil in the truck bed while the loading work to obtain the volume and the shape of loaded soil and to locate the position to load next. Especially, in order to measure the inside of the truck bed with high accuracy, we constructed a system to measure the loaded soil directly from a backhoe. We measure inside truck bed from multiple views while loading by mounting multiple sensors on the arm and on the cabin of the machinery. The performance of the proposed system is validated in an experiment with actual construction machineries at a test field.

Constructing lunar infrastructures including landing pad and habitable base is essential for enhancing lunar exploration and utilization in the future. The lunar surface regolith, fine-grained loose soil, needs to be compacted and consolidated enough for securely constructing the infrastructures. In this paper, we present a traction control of compaction roller towed by a small mobile robot. The control method aims to minimize the relative drawbar pull between the small mobile robot and the compaction roller. We also manufactured two different shapes of compaction rollers and verified the relationship between the roller shape and control performance. The experimental results show that the proposed traction control effectively minimizes the relative drawbar pull for the spiral-shaped roller while it becomes unstable for the cylindrical-shaped roller because of the roller excites soil bulldozing or excavation during its rolling.