This project aims to enable infrastructure construction in diverse environments using a collaborative AI robot system, with a focus on applications at natural disaster sites involving river channel blockages. To this end, we are advancing R&D in three areas: (1) System integration for emergency surveys of river channel blockages, (2) System integration for emergency restoration work of river channel blockages, and (3) Technologies for responding to river channel blockages. The second area—emergency restoration—requires work using construction machinery. However, since river channel blockages typically occur in mountainous regions, transporting large equipment is challenging. To address this, we aim to develop an emergency restoration system centered on the deployment of drainage pumps, using multiple compact construction robots (weighing less than 3 tons) that can be transported by helicopter. Figure 6 shows an overview of the system integration for emergency restoration. Below, we present the main outcomes of the 2024 efforts in the system integration of emergency restoration for river channel blockages.

In this R&D Item, we have been working on building an emergency restoration system by integrating technologies corresponding to the “Preparation for Emergency Restoration Work” and “Emergency Response” phases of the river channel blockage disaster response scenario shown in Figure 1. In river channel blockage disasters, there is a risk that natural dams formed by collapsed slopes may fail due to river overflow, potentially triggering large-scale debris flows. To prevent such secondary disasters, urgent countermeasures are required to redirect upstream water downstream. Therefore, in 2024, we focused on developing a system for deploying drainage pumps to transfer water from the upstream impounded area to the downstream. Since river blockages often occur in mountainous regions, transporting construction equipment is challenging due to the weight limits of helicopters. To address this, we developed a multi-robot system composed of compact construction robots that can be transported by helicopter without disassembly. This system enables emergency restoration work to be carried out by a small number of supervisors, without the need for on-site workers. Given the possibility of both gradual and sudden environmental changes during restoration in such settings, real-time monitoring is essential. To support this, we developed a stake-type sensor pods for collecting environmental data. These pods are installed using remotely operated construction robots capable of performing fine manipulation tasks. Figure 7 shows the insertion of a stake-type sensor pod into the ground using a construction robot in a simulated environment at Yanmar’s test site, as well as image data collected from the sensor installed in the Kyushu University field environment.

Next, we introduce the system for deploying drainage pumps. Compact construction machines, which can be transported without disassembly, allow for rapid response. However, during the process of submerging drainage pumps, there may be cases where the machine’s arm is not long enough to reach from the pond side. To address this, we developed a pump deployment system using a lightweight frame. Figure 8 illustrates the operation of the tool: by connecting two extension frames to the main frame holding the submersible pump, the left side is pushed forward, allowing the pump on the right side to be placed into the water. This system was tested via remote operation at a simulated river channel blockage embankment field owned by Kumagai Gumi’s research institute. Additionally, a crawler carrier dump truck was used to transport these tools, with some sections of travel conducted autonomously.