Development of 'Charging Amount Planning System' to Streamline Blasting Excavation in Mountain Tunnel Construction
Okumura Corporation, in collaboration with JMA Systems Corporation and Acroquest Technology Co., Ltd., has developed a system that uses AI to plan appropriate charging amounts for blasting excavation in mountain tunnel construction. The system aims to reduce reliance on skilled workers, stabilize quality, and improve productivity. The inference takes about 5 to 10 minutes, and field verification has confirmed its high accuracy.
📋 Article Processing Timeline
- 📰 Published: May 21, 2026 at 18:00
- 🔍 Collected: May 21, 2026 at 09:31
- 🤖 AI Analyzed: May 21, 2026 at 09:42 (10 min after Collected)
Okumura Corporation (Headquarters: Abeno-ku, Osaka, President: Takanori Okumura), in collaboration with JMA Systems Corporation (Headquarters: Minato-ku, Tokyo, President: Takeshi Sakakura) and Acroquest Technology Co., Ltd. (Headquarters: Kohoku-ku, Yokohama, President: Ryu Shinmen), has developed the 'Charging Amount Planning System', which uses AI to plan the appropriate charging amount for blasting excavation in mountain tunnel construction. This will lead to the stabilization of finished shape quality in blasting excavation without relying on skilled workers.
[Background]
In blasting excavation for mountain tunnel construction, skilled workers check the condition of the bedrock on the excavation face and plan the blasting pattern (drilling arrangement of charging holes and charging amount) based on their experience. The finished shape of the blasting excavation tends to depend heavily on the skill of the workers. On the other hand, addressing productivity improvement and the decline in the number of skilled workers has become an important issue in mountain tunnel construction. To tackle these issues, ICT construction technologies, such as computer jumbos, are being actively introduced to construction sites. We decided to develop a technology that can plan an appropriate blasting pattern according to the bedrock condition without relying on the skill of experienced workers by utilizing various data acquired through these ICT construction technologies.
[Overview and Features]
Based on drilling data acquired from a computer jumbo, this system uses an AI model trained on past construction data to plan the appropriate charging amount for each charging hole through three steps: '1. Drilling data acquisition and editing', '2. Charging amount inference execution', and '3. Inference result display and output'. It can plan the charging amount for the entire cross-section in a short time of about 5 to 10 minutes, from acquiring drilling data to displaying the inference results.
Overview of Charging Amount Planning System
1. Drilling data acquisition and editing
Transfer drilling data recorded in the computer jumbo to a dedicated PC, edit it as input data for the AI model, and then upload it to the cloud.
2. Charging amount inference execution
Calculate the planned charging amount for each charging hole using the AI model on the cloud, considering the number of detonator stages (initiation timing).
3. Display and output of inference results
Display the calculated planned charging amount graphically.
Display example of planned charging amount
[Field Verification]
We verified the inference results of the charging amount by this system in the 'FY2023-2025 Kubokawa Saga Road Mitsuke Tunnel Construction' ordered by the Shikoku Regional Development Bureau.
Targeting the outer peripheral charging holes, we compared the planned charging amount per charging hole by this system with the actual value determined by a skilled worker. In the cross-section of the blasting excavation where the excavation volume was close to the designed cross-section, the planned value was 340-390g compared to the actual value of 400g per charging hole. From this result, we confirmed that the charging amount calculated by this system is generally consistent with the actual value and that appropriate charging amounts can be planned.
Charging hole drilling situation by computer jumbo
[Future Development]
In the future, we will actively deploy this system to tunnel sites using computer jumbos, and strive to advance this system by working on accuracy verification and improvement of the AI model. At the same time, we will also work on developing a system to plan the drilling arrangement of charging holes, aiming to develop it into an advanced blasting pattern planning system integrated with the charging amount planning system.
[Inquiries Regarding This Matter]
Okumura Corporation
Technology Division, Technology Research Institute, Planning and Management Group
TEL: 029-865-1521
FAX: 029-865-1522
E-mail: giken@okumuragumi.jp
[Developer in Charge]
Okumura Corporation
Technology Division, Technology Research Institute, Civil Engineering Research Group
Hajime Hamada
[Background]
In blasting excavation for mountain tunnel construction, skilled workers check the condition of the bedrock on the excavation face and plan the blasting pattern (drilling arrangement of charging holes and charging amount) based on their experience. The finished shape of the blasting excavation tends to depend heavily on the skill of the workers. On the other hand, addressing productivity improvement and the decline in the number of skilled workers has become an important issue in mountain tunnel construction. To tackle these issues, ICT construction technologies, such as computer jumbos, are being actively introduced to construction sites. We decided to develop a technology that can plan an appropriate blasting pattern according to the bedrock condition without relying on the skill of experienced workers by utilizing various data acquired through these ICT construction technologies.
[Overview and Features]
Based on drilling data acquired from a computer jumbo, this system uses an AI model trained on past construction data to plan the appropriate charging amount for each charging hole through three steps: '1. Drilling data acquisition and editing', '2. Charging amount inference execution', and '3. Inference result display and output'. It can plan the charging amount for the entire cross-section in a short time of about 5 to 10 minutes, from acquiring drilling data to displaying the inference results.
Overview of Charging Amount Planning System
1. Drilling data acquisition and editing
Transfer drilling data recorded in the computer jumbo to a dedicated PC, edit it as input data for the AI model, and then upload it to the cloud.
2. Charging amount inference execution
Calculate the planned charging amount for each charging hole using the AI model on the cloud, considering the number of detonator stages (initiation timing).
3. Display and output of inference results
Display the calculated planned charging amount graphically.
Display example of planned charging amount
[Field Verification]
We verified the inference results of the charging amount by this system in the 'FY2023-2025 Kubokawa Saga Road Mitsuke Tunnel Construction' ordered by the Shikoku Regional Development Bureau.
Targeting the outer peripheral charging holes, we compared the planned charging amount per charging hole by this system with the actual value determined by a skilled worker. In the cross-section of the blasting excavation where the excavation volume was close to the designed cross-section, the planned value was 340-390g compared to the actual value of 400g per charging hole. From this result, we confirmed that the charging amount calculated by this system is generally consistent with the actual value and that appropriate charging amounts can be planned.
Charging hole drilling situation by computer jumbo
[Future Development]
In the future, we will actively deploy this system to tunnel sites using computer jumbos, and strive to advance this system by working on accuracy verification and improvement of the AI model. At the same time, we will also work on developing a system to plan the drilling arrangement of charging holes, aiming to develop it into an advanced blasting pattern planning system integrated with the charging amount planning system.
[Inquiries Regarding This Matter]
Okumura Corporation
Technology Division, Technology Research Institute, Planning and Management Group
TEL: 029-865-1521
FAX: 029-865-1522
E-mail: giken@okumuragumi.jp
[Developer in Charge]
Okumura Corporation
Technology Division, Technology Research Institute, Civil Engineering Research Group
Hajime Hamada
FAQ
How many minutes does it take for the charging amount planning system to create a plan?
It can plan the charging amount for the entire cross-section in about 5-10 minutes, from acquiring drilling data to displaying inference results.
What data does this system use for inference?
It infers based on drilling data acquired from computer jumbos and an AI model trained on past construction data.
What was the difference between the planned and actual values in the field verification?
The planned values were 340-390g compared to the actual value of 400g, confirming that it aligns well and enables appropriate planning.