AI Construction Near Miss Detection Enhances Safety

A Japanese technology collaboration between Sumitomo Heavy Industries and NEC Corporation has developed an advanced artificial intelligence system designed to automatically detect and report near-miss incidents on construction sites.AI construction near miss detection

The companies say the system is a first-of-its-kind solution capable of automatically extracting, visualizing, and summarizing potentially dangerous situations using camera footage and operational data from construction equipment.

AI System for Construction Safety

The new technology analyzes video footage and sensor data captured from hydraulic excavators, identifying potentially hazardous situations that could lead to accidents.

Using advanced AI capabilities, the system automatically generates reports based on incidents detected during construction activities. The goal is to help companies improve safety by identifying risks before accidents occur.

The project combines the engineering expertise of Sumitomo Heavy Industries with the artificial intelligence and video analysis technology developed by NEC Corporation.

How the Technology Works

The system operates through a multi-step process:

  • Video capture: Cameras and sensors installed on hydraulic excavators record site operations.
  • Risk detection: AI analyzes the footage to identify potential safety risks or unsafe behaviors.
  • Cross-referencing: Hazardous situations are compared with work logs and operational data.
  • Automatic reporting: The system generates reports highlighting near-miss incidents and equipment-related risks.

The technology uses NEC’s proprietary video recognition tools combined with generative AI, enabling the system to analyze complex construction environments and summarize safety concerns.

Identifying Hazardous Behaviors

The AI platform is designed to recognize several potential risks on construction sites, including:

  • Unsafe worker behavior near heavy equipment
  • Potential equipment malfunctions or operational failures
  • Situations that could lead to accidents involving machinery

By automatically detecting and reporting these situations, construction managers can review incidents quickly and implement preventive measures.

Development Timeline

The joint team successfully created a proof of concept in September 2025, demonstrating the system’s ability to identify safety risks from recorded footage.

The companies plan to move toward practical implementation by 2027, aiming to deploy the technology across construction projects where safety monitoring is critical.

Future Applications

Beyond identifying near-miss incidents, the developers plan to expand the system’s capabilities to detect unsafe working conditions that may otherwise go unnoticed.

Future versions may also evaluate site operations against specific safety rules and operational guidelines, allowing construction companies to ensure better compliance and improve workplace safety standards.

If successfully implemented, the AI system developed by Sumitomo Heavy Industries and NEC Corporation could significantly enhance safety monitoring across construction sites by providing real-time insights into potential hazards.

Latest from Blog

African Mining Week 2026 Energy-Mining Nexus Insights

African Mining Week 2026 Energy-Mining Nexus Insights

African Mining Week 2026 will spotlight the growing energy-mining nexus as African countries accelerate industrial development and expand critical mineral production across the continent. The conference, taking place October 14–16, 2026, in
Botswana Cloud and Cyber Security Services Explained

Botswana Cloud and Cyber Security Services Explained

Liquid Intelligent Technologies has expanded its cloud and cyber security services in Botswana, strengthening the country’s digital infrastructure and supporting national digital resilience initiatives. The announcement coincides with the company’s 10-year anniversary
Liebherr Environmental Award Recognizes Innovation

Liebherr Environmental Award Recognizes Innovation

Liebherr USA, Co. mobile and crawler cranes received the Liebherr environmental award from the Specialized Carriers & Rigging Association (SC&RA) during the organization’s Annual Conference held on Amelia Island in April 2026.
Hexagon GeoCloud Webinar: Transforming Data Management

Hexagon GeoCloud Webinar: Transforming Data Management

ThThe upcoming Hexagon GeoCloud webinar will show how the company’s new cloud platform combines survey and reality capture workflows into one connected system. As digital projects become more data-driven, many organizations now
Malta Dynamics Joins Vertical Supply Group Today

Malta Dynamics Joins Vertical Supply Group Today

Malta Dynamics has officially joined Vertical Supply Group, marking a strategic move aimed at strengthening safety solutions across the construction and industrial sectors. The partnership combines Malta Dynamics’ expertise in fall protection
RAP Technology in Asphalt Production Explained

RAP Technology in Asphalt Production Explained

RAs sustainability and resource efficiency become increasingly important in global infrastructure projects, Reclaimed AsphaltAs sustainability and resource efficiency become more important in global infrastructure projects, Reclaimed Asphalt Pavement (RAP) is becoming a
CNH 2026 AGM Voting Results: Key Highlights

CNH 2026 AGM Voting Results: Key Highlights

CNH Industrial N.V. has announced the voting results of its 2026 Annual General Meeting (AGM) while also releasing its 2025 Sustainability Report. The company confirmed key board appointments, approved its annual financial
Laser Scanning in Structural Engineering Explained

Laser Scanning in Structural Engineering Explained

Laser scanning in structural engineering is becoming an essential tool for evaluating existing buildings, planning repairs, and preserving historic structures. As projects become more complex, reality capture technologies are helping engineers improve