Chinese engineers wire Raspberry Pi into 600-meter railway tunnel to find any holes

Chinese scientists have found a novel use for the Raspberry Pi: detecting voids in the linings of railway tunnels that could lead to structural damage, or even collapse.

News of the Pi's potential new role appeared last week in the journal Buildings – the organ of the International Council for Research and Innovation in Building and Construction – in a paper penned by authors from Gansu Road & Bridge Construction Group Co, and the School of Civil and Hydraulic Engineering, Huazhong University of Science and Technology.

The paper, titled Raspberry Pi-Based IoT System for Grouting Void Detection in Tunnel Construction, opens by explaining that railways tunnels are built with primary and secondary tunnel linings. The latter enhances a tunnel's strength and ability to handle local stresses. But if secondary linings contain voids – which can form either because of shoddy work or shifts in geological conditions – it's possible the integrity of a tunnel could be compromised.

At this point The Register imagines all readers will agree that a system to detect voids before they become a problem would be a fine idea.

That's what the authors wanted, too. But their assessment of techniques like manual inspection (tapping walls to listen for odd sounds) or using ground penetrating radar found both had limitations. Ground penetrating radar, specifically, is very accurate but costly and does not provide continuous monitoring.

Electronic sensors have already been used in other tunnels – to measure stress, humidity, temperature, and other variables.

But the authors of this paper wanted to test something different: the conductivity of concrete. Because when that dips it's a strong indicator of possible tunnel lining voids. The Raspberry Pi's 40 General Purpose Input Output (GPIO) pins make the small machines ideal for that sort of role.

• Raspberry Pi AI Camera takes inferencing load off the CPU
• Post-IPO Raspberry Pi results in: So you can make money in tech without added AI
• Hey kids, wanna clock the ISS? ESA's Astro Pi challenge runs your code in space
• Raspberry Pi 5 slims down for cut-price 2 GB RAM version

The authors embedded wires in the secondary lining of a recently built 600 meter tunnel, and hooked them up to Pin 2 as a source of current and Pin 34 to ground the circuit.

"If the circuit remains open (indicating an empty void), the Raspberry Pi detects no current flow. When the circuit is closed (indicating a filled void), the current flows through, and the Raspberry Pi registers this as a completed grouting process," the paper explains. The rig also used a Honeywell pressure transducer, and a temperature–humidity sensor.

The Raspberry Pi is not regarded as one of the world's more physically resilient computers, so the researchers used waterproof enclosures to protect the devices from the underground environment.

Readings from all the sensors were uploaded to Alibaba Cloud over a 5G connection, and stored in a MySQL database that "processes the incoming data, aggregating and organizing them into usable formats for real-time monitoring and historical analysis."

The authors suggest machine learning could be applied to the collected sensor data, or even to build digital twins of tunnels and allow real-time decision support.

Sending data to the cloud also means it can be accessed remotely: engineers don't need to go underground to assess a tunnel.

The authors validated results gathered using the rPis by comparing against data from ground-penetrating radar, and found that aside from a few limitations – the system did struggle to detect voids of irregular shapes or very small sizes – the mighty mini-machines did a very decent job. ®