On-site and virtual spaces are interconnected
Realization of the "digital twin"
Arayais aimed at an advanced future of construction where the real world and virtual space (simulator) are linked and integrated.
The simulator reflects on-site conditions and feeds back automatic control instructions based on analysis results from the simulator to the site, allowing heavy equipment and robots to operate automatically.
This contributes to the construction industry by improving productivity, reducing manpower, and enhancing safety.
BIM/CIM LINKED CRANE SIMULATOR
Enables crane planning studies in 3D using BIM/CIM data.
The system makes studies more efficient through collision determination, lifting capacity determination, and animated images of lifting and transporting.
It can be used not only for pre-construction planning studies, but also for plan reviews during construction, facilitating information sharing and fostering common understanding among the parties involved, thereby improving the efficiency and sophistication of construction work.
Flow of using the crane simulator
This simulator can be installed on the user's computer and used in the user's environment.
Users can input their BIM/CIM data and point cloud data, and use each function to perform planning studies on the Unity simulator.
As a result of the study, 3D movies (animations) and various information can be output.
*Please feel free to contact us for a detailed explanation and procedures.
Crane Simulator Demo Movie
Various conditions can be checked and paths can be studied on the simulator, including dynamic simulation.
Examples of Functions
- Crane model with hoisting/swiveling/upsetting motion
- Route Confirmation
- Collision Confirmation
- Automatic load rating calculation
- Setting of no-entry areas
- 3D VIDEO SIMULATION
- Reflection of in-situ point cloud data
JFE ENGINEERING'S BIM/CIM LINKED CRANE SIMULATOR FROM ARAYAWAS INTRODUCED FOR PLANNING STUDIES AT A BRIDGE SITE.
Efforts to improve productivity in the construction industry are accelerating due to such factors as the enforcement of laws related to work style reform.
JFE Engineering Corporation has been one of the first in the industry to promote on-site reform through DX using ICT and advanced technologies.
In this context, the crane simulator provided by Araya, which can be linked to BIM/CIM, has been introduced to JFE Engineering for planning studies at its bridge sites.
Automate crane planning studies (patented)
We are developing a system that can optimize crane plans and propose plans using "BIM x simulator (Unity) x optimization algorithm".
Case Study] Construction Order Optimization of Parts
The optimization algorithm was used to construct a two-story building on the simulation (with a crane to transport the components).
Faster construction could be achieved by increasing the amount of computation (number of optimization cycles) in the optimization algorithm.
HYDRAULIC EXCAVATOR AI AUTOMATION
The construction industry has long been in need of development of unmanned and automated technologies due to a growing shortage of workers, including the retirement of skilled technicians due to the aging of the workforce, as well as the danger of major accidents such as building collapses and mudslides occurring at work sites.
ArayaAt MHI, we are promoting the automation of hydraulic excavators using AI technology.
This will contribute to labor saving, efficiency, and safety improvement of work.
Reinforcement learning AI and imitation learning AI are used to help automate
Reinforcement learning" involves the AI itself performing various actions in a simulated environment through trial and error, learning and acquiring better actions to achieve its goals.
Although it takes more time to learn, it has the potential to acquire more operations than humans.
In "imitation learning," the AI learns similar actions based on the model data of a human (skilled engineer).
In a relatively short time, it is possible to achieve the same level of movement as humans.
THE AI ACQUIRES THE MANEUVERING METHOD THROUGH TRIAL AND ERROR.
AI ACQUIRES THE METHOD OF MANEUVERING BY IMITATING HUMAN EXAMPLES.
AI automation of hydraulic excavators in the real world
- IN THE REAL WORLD, AI (OFFLINE REINFORCEMENT LEARNING) AUTOMATES THE ENTIRE PROCESS FROM EXCAVATION OF FILL TO LOADING ONTO DUMP TRUCKS!
- What is off-line reinforcement learning?
- Learning is possible without a simulator if there is a significant amount of actual operation data. No need to adjust the gap and no need to reflect the simulator results in the real world.
- Autonomous AI (reinforcement learning) can autonomously learn superior action policies from actual operation data Remove inappropriate/skillful data from data
Utilizing point cloud data
SLAM (Self Location Awareness)
Using point cloud data acquired from LiDAR sensors, etc., it is possible to determine the self-position of robots and heavy equipment, as well as the surrounding environment.
This technology is especially useful indoors and in tunnels where GPS cannot reach.
SELF-POSITIONING BY SLAM
Using LiDAR and other technologies, the system supports autonomous operation of construction machinery/heavy equipment through spatial recognition (self-positioning, understanding of the site).
SELF-POSITIONING IS POSSIBLE EVEN IN ENVIRONMENTS WHERE GPS IS NOT AVAILABLE.
Stereo camera (Depth camera)
Automated operation of equipment for measurement with Nishimatsu Construction Co.
Began automation of construction equipment in mountain tunnel construction
Ltd. and Nishimatsu Construction Co., Ltd. have jointly realized the automated operation of the measuring device "Tunnel RemOS-Meas.
This time, the automation technology of Arayawas incorporated into the measuring device, which is one of the systems for building remote and automated technology for various construction machines, in Nishimatsu's efforts to unmanned mountain tunneling face work.
Infrastructure and plant inspection
In recent years, inspections have become increasingly important as various infrastructure structures have aged.
However, inspections by humans are time-consuming, labor-intensive, and sometimes dangerous, so there is a need to improve inspection efficiency and reduce manpower.
Arayais providing a solution that uses image recognition AI to detect rust, cracks, and other problems.
RUST/CORROSION DETECTION AI
Rust and corrosion on bridges and steel towers are detected by image recognition AI.
It is also possible to rank inspection results according to the degree and area of rusting.
CRACK DETECTION AI
Cracks in concrete and other materials are detected by image recognition AI.
Cracks as small as 0.2 mm could be detected.
Air conditioning optimization solutions
Air conditioning is said to account for approximately 48% of a building's energy consumption.
ArayaAt Rinnai Systems, we provide AI air conditioning optimization solutions for district heating and cooling (DHC), large heat sources, and building central air conditioning, helping to reduce electricity consumption, electricity rates, and labor costs by automating tasks that were previously performed by skilled workers.
Crane part information can be assigned to any 3D model, such as slewing body, boom, and hook, greatly increasing the number of crane types that can be used as 3D dynamic models.
In addition, crane movements can be managed parametrically from bridge girder ground clearance to erection, making it easy to edit the crane movements as they change from time to time.
BIM/CIM Linked Crane Simulator
HYDRAULIC EXCAVATOR AI AUTOMATION DEMO VIDEO
Using the industrial simulator "Vortex Studio" (provided by CM Labs Simulations and Dentsu International Information Service), a hydraulic excavator was trained to dig a certain amount of soil by reinforcement learning.
At first, the excavator could not scoop soil at all, but it gradually learned the angle of the bucket and the timing of its movement, and was able to scoop more than a certain amount of soil.
What is SLAM (Simultaneous Localization and Mapping)?
SLAM (Self-Localization Awareness) is a generic term for technologies that enable robots and other mobile objects to simultaneously identify their own location and map their environment.
One area where SLAM is useful is in automated driving and autonomous moving systems.
When constructing an automated driving system, there are three main categories: "recognition," "judgment," and "operation.
Specifically, the system acquires point cloud data from LiDAR sensors, etc., and uses the map information and self-positioning constructed by SLAM to perform specific tasks, such as avoiding obstacles, for example.
What is BIM/CIM (Building/Construction Information Modeling)?
BIM/CIM is intended to facilitate the sharing of information among the parties concerned throughout the project by introducing 3D models from the planning, survey, and design stages, and by linking and developing 3D models in the subsequent construction and maintenance stages, thereby making a series of construction production and management systems more efficient and advanced. The project aims to facilitate the sharing of information among the parties involved throughout the project and to improve the efficiency and sophistication of construction production and management systems.
By utilizing the latest ICT to share information in the planning, investigation, design, construction, and management phases of the construction production system, an efficient and high-quality construction production and management system will be established.
This is expected to significantly reduce errors and rework, reduce the number of simple tasks, improve safety at construction sites by shortening processes, improve business efficiency and economic benefits, and, as a side effect, improve people's lives through better infrastructure development and maintenance, and increase the motivation of people engaged in the construction industry and enhance their sense of fulfillment. In addition, it is expected to improve people's lives through better infrastructure development, maintenance, and management.
(From Ministry of Land, Infrastructure, Transport and Tourism: BIM/CIM Portal Site)