From October 15 to 17, a full-scale fire experiment was conducted in an actual tunnel environment at the No. 1 shaft worksite of the Gaoligong Mountain Tunnel on the Dali–Ruili Railway in Yunnan. Professor Maohua Zhong’s research group at Tsinghua University’s School of Safety Science joined Chongqing University, China Railway Tunnel Group and JX Intelligence to investigate fire-induced smoke spread and the coordination of fixed and mobile monitoring systems.

When a tunnel fire breaks out, flames are only part of the concern. Where will the smoke spread? How will temperatures change? Where will hazardous gases accumulate? These invisible, moving and constantly evolving hazards are central to both fire research and emergency response.

Laboratory simulations help researchers understand the underlying mechanisms. However, the spatial scale, environmental conditions and smoke propagation processes of an actual tunnel also require on-site observation and recording. For the participating teams, the experiment provided an opportunity to test equipment, collect real-world data and explore monitoring methods in a complex engineering environment. JX Intelligence supplied robots and environmental sensing equipment and supported on-site data acquisition.

Taking Research from the Laboratory into an Actual Tunnel

Tunnels are long, narrow and relatively enclosed. During a fire, smoke spread, temperature changes and declining visibility interact with one another. Understanding these processes requires researchers to establish what happened, when and where it occurred, and how conditions evolved.

Laboratory simulations are essential research tools, but full-scale tunnels present more complex spatial configurations and environmental conditions.

At the Gaoligong Mountain Tunnel, the research team brought testing into an actual engineering setting. The experiment focused on smoke propagation, collecting measurements such as smoke temperature and recording how smoke spread through the tunnel. These observations provide first-hand data for studying tunnel fire development under real-world conditions.

Continuous, accurate data recording is essential to reconstructing changes in on-site conditions during subsequent analysis.

Measurements from a single location describe conditions at that point, while hazards within a tunnel evolve across both time and space. One focus of the joint experiment was therefore to explore how measurements from different locations could supplement fixed monitoring.

Complementing Fixed Monitoring with Mobile Sensing

Fixed sensors continuously observe specific locations, providing consistent data over time. Mobile platforms can change position according to task requirements, extending spatial coverage.

Each approach offers distinct advantages. The experiment explored coordinated fixed and mobile monitoring, examining how stationary sensors and robot-based mobile sensing could complement one another to provide a more comprehensive picture of tunnel conditions.

JX Intelligence supplied two TitanRanger X1U robots, together with heat flux, temperature and gas sensing equipment.

Heat flux sensors measured heat flux density; temperature sensors and thermocouples collected temperature data; and gas sensors collected data on hazardous gases. Together, the robotic platforms and multiple sensor types supported mobile sensing and multi-source data acquisition during the experiment.

For on-site data acquisition, sensor readings are only the starting point. Without corresponding timestamps and spatial locations, temperature, heat flux and gas concentration measurements are difficult to relate to smoke propagation or compare across locations and stages of the experiment.

To meet the experiment’s data requirements, JX Intelligence developed algorithms that enabled the backend system to export raw data with timestamps and spatial coordinates (X, Y, Z).

Timestamps allow researchers to track how measurements change as the experiment progresses. Spatial coordinates link those measurements to specific locations within the tunnel.

This provides a basis for organizing and analyzing data from multiple sources, supporting further investigation into smoke spread, environmental changes and the relationship between different monitoring approaches.

Turning a Field Experiment into Data for Analysis

The value of a full-scale field experiment extends beyond completing the test itself. It also lies in preserving a record of processes as they unfold under real-world conditions.

Fire-generated smoke does not remain at a fixed location. It evolves over time and moves through space. Field data allows researchers to examine this continuous process, analyze environmental changes at different locations and develop a deeper understanding of smoke propagation.

The data collected during this experiment can inform tunnel fire research and support further investigation into hazard monitoring and emergency response.

A more complete understanding of how hazards emerge and develop provides a stronger basis for research into prevention and response.

Located in a complex mountainous environment, the Gaoligong Mountain Tunnel requires sustained attention to safety during both construction and operation. Conducting research in an actual tunnel means working directly with site conditions. It also puts equipment, sensors, data processing methods and coordination among participating teams to the test.

Tsinghua University, Chongqing University, China Railway Tunnel Group and JX Intelligence contributed their respective research and technical expertise to complete the joint experiment.

For JX Intelligence, the experiment also provided practical experience in applying its robotic platforms and environmental sensing capabilities in confined spaces. TitanRanger X1U robots, multiple types of environmental sensors and data processing algorithms worked together to support on-site data acquisition, connecting equipment capabilities with research requirements in an actual engineering environment.

From sensor readings inside the tunnel to raw data tagged with time and location, and ultimately to researchers’ analysis of smoke propagation, each stage helps preserve a record of real-world conditions. These records provide a foundation for subsequent research and practical reference data for studies of tunnel fire prevention and emergency response.

JX Intelligence thanks Professor Maohua Zhong’s research group at Tsinghua University’s School of Safety Science, Chongqing University, China Railway Tunnel Group and the on-site personnel for their collaboration. The company will continue to contribute its robotic platforms and environmental sensing technologies to joint testing and the exploration of potential applications in complex environments, supporting safety monitoring and emergency detection in tunnels and other confined spaces.