Idler Condition-Based Monitoring Systems: Review

Mining industries heavily rely on conveyor belt systems to transport bulk materials efficiently and safely. Conveyor idlers (rollers) are critical components in overland and underground mining conveyors, and their failures are a leading cause of costly downtime [6]. Idler roller failures can lead to multiple detrimental outcomes, such as belt damage, reduced conveyor efficiency, increased energy consumption, elevated repair costs, and substantial downtime. Failed idlers can cause misalignment of conveyor belts, increasing friction and wear, ultimately resulting in severe damage or even fire hazards. With thousands of idlers per conveyor (up to ~2,700 per km of belt [4]), traditional maintenance methods involved periodic manual inspections, which are labour-intensive, expensive, unreliable, and pose safety risks. The direct financial impact includes production losses, repair and replacement costs, and labour expenses associated with urgent corrective maintenance. This has driven development of a range of condition-based monitoring (CBM) technologies to detect failing idlers early and enable predictive maintenance, facilitated by advanced technologies. Enabling predictive maintenance offers significant improvements in safety, reliability, and cost efficiency.

The primary objective of this research is a thorough comparative evaluation of available idler monitoring technologies. Below, we survey all current and emerging idler CBM solutions – how each works, key components, and typical use cases.

Traditional idler maintenance has been predominantly reactive or preventive. Reactive maintenance responds only after equipment failures occur, resulting in prolonged downtimes and higher costs. Preventive maintenance, while scheduled regularly, does not prevent unexpected failures. The emergence of predictive maintenance technologies – such as vibration analysis, temperature monitoring, acoustic emission detection, and wireless sensor networks – offers significant improvements by allowing the early detection of impending failures. By integrating online condition monitoring systems with smart sensors and robotics, operators can proactively manage idler roller health, eliminating single points of failure in material handling. This technology reduces inspection times drastically, enables real-time problem detection, and optimises maintenance schedules, ultimately improving mine operational efficiency, thus minimising downtime and optimising resource allocation.

One approach is to attach sensors on idler frame brackets to measure vibration or noise signatures of the rollers. For example, the Intium Roller Condition Monitoring (RCM) system uses small battery-powered accelerometer nodes mounted on every idler frame (carrying and return). These sensors form a wireless mesh network to a local collector hub, which uploads data to a cloud platform for analysis. Vibration levels are reported frequently (e.g. every 2 minutes) and the system’s software flags abnormal signatures, providing a prioritised list of idler frames likely exhibiting faults [24]. Such frame-mounted sensor systems can often be retrofitted without modifying the rollers themselves. They typically run on replaceable batteries (RCM uses AA cells) and use low-power wireless (e.g. proprietary mesh or Wi-Fi) to relay data. These systems are commercially available and have been trialled in industry, proving capable of detecting failures along entire conveyors with high spatial resolution and predictive accuracy [7]. A limitation is that the sensor is not on the roller bearing directly, so very subtle incipient faults might be masked by background belt vibration. However, in practice they have shown effective early fault warnings – one study noted such systems could detect idler faults up to ~280 hours (11+ days) earlier than manual inspection [24]. Use cases include overland conveyors where continuous remote monitoring is needed to avoid sending crews to hard-to-access idler locations.

Many idler failures, especially bearing failures, manifest as excess heat due to friction. Thus, temperature-based monitoring is a popular technique to catch rollers that are running hot (a precursor to seizure or fire) [8]. One implementation, mentioned below, is embedding temperature sensors directly at the bearing (e.g. the Smart-Idler) to continuously measure bearing temperature. These sensors can alert when a bearing’s temperature exceeds a threshold above the set temperature, indicating likely internal damage.

A recent addition to this category is a IoT-based monitoring system that uses an infrared temperature sensor, a thermal imaging sensor, and a visual camera to continuously track individual idler temperatures. Developed as an open-source solution, the system logs and transmits data via a Raspberry Pi and cellular or Wi-Fi connectivity, issuing real-time alerts if a preset threshold is exceeded. A companion web dashboard and Telegram bot interface allow remote access to thermal maps, photos, and video clips, offering a flexible and scalable tool for targeted fire prevention. While still in the prototype stage, this system has been successfully tested in field conditions and received strong user acceptance [1].

Thermal monitoring is especially crucial in underground coal conveyors, where an overheated idler can ignite coal dust or the belt; regulations often mandate temperature sensors or fire detection cables for early warning. In summary, temperature-based methods are very effective at detecting the late-stage fault condition (overheating) and are straightforward to interpret. However, they typically only catch failures in relatively advanced stages (when heat is generated), which might be after significant bearing damage has occurred. Thus, they are excellent for fire prevention and maintenance scheduling a few hours or days ahead, but not as sensitive for long-term predictive prognostics.

Failing bearings and rollers emit distinctive sound signatures – from subtle ultrasonic scrapes to loud squeals or rattling noises. Acoustic-based idler monitoring uses microphones or acoustic sensors to detect these sound anomalies. In maintenance practice, handheld ultrasonic “listening” devices (like the SKF Idler Sound Monitor kit) have been used: a technician points an ultrasonic microphone toward idlers and listens for the tell-tale high-frequency noise of a bad bearing [30]. Modern CBM solutions aim to automate acoustic detection. One challenge is that conveyors themselves are noisy, and distinguishing a failing idler’s sound can be difficult without close proximity or advanced filtering [19].

A newer strategy embeds sensors inside the idler rollers themselves. The Vayeron Smart-Idler® is a leading example – a drop-in replacement roller that contains an internal sensor module measuring bearing vibration (envelope analysis), RPM, and temperature [12], [13]. The module is completely wireless and self-powered via an internal energy-harvesting generator driven by the roller’s rotation. As the roller turns, the Smart-Idler continuously monitors the health of the bearing and shell – detecting increases in vibration or noise indicative of spalling, or rising temperature caused by friction . Data is sent via a wireless mesh network, with antennas discreetly integrated at the roller shaft, providing a real-time 24/7 view of all rollers’ condition. Smart-Idler’s analytics perform spectral analysis on the sensor signals to accurately identify bearing faults even in the noisy conveyor environment [3], [12], [13]. In tests it has been able to detect developing roller failures up to 9 months in advance of a breakdown [2], allowing very early intervention. Components of such systems typically include the instrumented idler – integrating a built-in sensor module (includes processor, radio, and energy harvester) – alongside wireless gateways. Smart idlers are particularly suited for operators requiring continuous condition monitoring of each roller. These systems improve safety by reducing the need for personnel to inspect belts in dangerous or remote locations, and they help prevent catastrophic failures or fires by detecting issues like bearing wear or overheating at early stages. Smart idler technologies are being adopted in both surface and underground mines, with features like sealed design and absence of external wiring making them intrinsically safe, which is critical for compliance in underground coal mining environments [26].

Several companies are exploring similar “sensing idler” concepts. For instance, Artur Küpper GmbH has developed a prototype with embedded digital temperature sensors placed in the axle near each bearing, networked together for real-time condition monitoring. Their design uses two temperature sensors per idler and transmits data through the idler station via wired contacts to a local monitor [10]. This emphasises bearing temperature as a simple, low-cost indicator of failure (avoiding more expensive vibration or infrared systems).

An emerging high-tech solution uses fibre optic cable as a continuous listening device along the conveyor. In Mining3’s research (now commercialised as AURA IQ), a fibre optic cable is strapped or clamped to the conveyor’s steel structure for the entire length [11]. The fibre is connected to an interrogator unit at one end that sends rapid laser pulses down the fibre and measures the backscattered light (using Rayleigh backscatter). Minute vibrations in the conveyor structure (caused by idler bearings grinding or belts flapping) disturb the fibre and produce measurable changes in the backscatter signal. By analysing the timing and frequency of these changes, the system can determine where along the belt a particular acoustic pattern originates and what its characteristics are [15]. Essentially, the fibre serves as thousands of microphones arrayed every few meters. Importantly, these systems perform frequency-domain analysis to isolate signature frequencies of bad bearings (as opposed to random noise). The entire conveyor can be monitored simultaneously by one interrogator, with typical coverage up to ~10 km of belt per unit. Components: a fibre optic cable (often standard telecom single-mode fibre) running along the conveyor structure, and an interrogator/data acquisition box. No power or electronics are needed in the field – the fibre is passive [16]. This makes DAS very attractive for hazardous locations (e.g. underground coal) because there are no ignition sources along the belt.

Another fibre-optic approach specifically monitors temperature rather than vibration. DTS systems (offered by companies like Yokogawa, AP Sensing, etc.) use optical fibres with the Raman scattering principle to measure temperature along their length [9]. When laser pulses are sent through the fibre, a small fraction of light is scattered back with a temperature-dependent spectrum (the ratio of certain wavelength components varies with temperature). By analysing this backscatter, the DTS interrogator can read the temperature at every point along the fibre, effectively providing a thermometer spanning the whole conveyor [9]. In conveyor applications, a DTS cable may be run along the conveyor’s idler frames or even woven through the idler hollow shafts (in research settings) to directly sense heat at the bearings. More commonly, fibre-based heat detection is used for fire detection: the fibre is routed above the belt so that if a smouldering roller or friction heats the area, the fibre picks it up. Commercial DTS units can take temperature readings at high spatial resolution (e.g. every 1 m) and update frequently (every few minutes or faster) [9]. One reported deployment monitored each 1 m segment of conveyor once per day, but systems can be configured for much higher frequency if needed.

However, because DTS only measures temperature, its detection capability is limited to late-stage failures (when significant heating has occurred). It may miss early fault indicators that have not yet produced heat. Some operations use DTS in combination with other methods: the DTS gives a safety overlay for catching anything that overheats, while vibration/acoustic sensors give predictive warning earlier.

Summary of Fibre Optic Systems:

Fibre solutions shine in environments where electronics maintenance is problematic (long, remote conveyors or gassy underground mines) – once installed, the fibre cable requires virtually no maintenance, but calibration [15]. Both DAS and DTS provide continuous coverage over long distances at relatively low incremental cost per length. They also inherently pinpoint the location of a fault (by the time-of-flight of the signal changes) along the belt to within a meter or two [16]. The high sensitivity of acoustic fibre means even minor changes in sound/vibration can be caught, but this also means advanced data analytics are essential to filter out normal operational noise. Overall, fibre-optic monitoring is considered a cutting-edge and high-potential emerging technology in conveyor CBM, with successful trials showing it can detect idler failures that were previously only found by laborious walking inspections [25], [16].

This is targeted at large, critical conveyors (mainly overland) where even brief downtime is extremely costly, and where automated, continuous idler replacement can save hours of shutdown. It’s been field-tested at industrial scale [22]. Drawbacks include the need to install rails along the conveyor (an infrastructure investment) and the fact that it addresses failures after they occur (though very rapidly). Spidler may be less suitable underground, especially coal mines, due to the diesel power and space requirements.

Another approach keeps the human out of the equation but still focuses on inspection rather than replacement. In this category, a robot (or vehicle-mounted system) travels along the conveyor to examine idler conditions using sensors like cameras or lasers. The Robotic Idler Predict system is one example: it’s essentially a robot arm on a mining vehicle (such as a pickup or dedicated trolley) that can be driven alongside a conveyor. The 6-axis robotic arm automatically positions a sensor head under each idler along a loaded conveyor to check its condition [6].

UAVs equipped with cameras or sensors represent a flexible, non-contact way to survey conveyor idlers. Drones can fly over or alongside overland conveyor segments that are difficult to access on foot (e.g. across gullies or in mountainous terrain). They can carry thermal cameras to detect hot idlers and high-resolution video to possibly spot visible signs of failure (like a stuck or misaligned roller). The concept is that a drone could autonomously fly the length of a conveyor once per day (or as needed), capturing imagery of every idler set [23]. Some advanced drone systems might also use acoustic sensors, hovering near the conveyor to “listen” for bad bearings. The advantage of UAVs is rapid coverage of even the longest conveyors – a drone can inspect tens of kilometres in a single flight, which would take humans many hours. They also don’t require any fixed installation on the conveyor.

Drones are being tested for routine conveyor inspection at large open-pit mines and industrial sites. For example, a drone could fly each morning along a 5 km conveyor and feed thermal images into an algorithm to automatically flag hotspots. This turns a formerly laborious task into a quick remote operation. However, there are significant challenges before UAVs become a mainstream CBM tool:

• Flight stability and navigation in the conveyor environment – GPS may not be reliable near structures or underground, and there can be high winds or dust.
• Battery life – drones have limited endurance (20–30 minutes typical), so multiple units or battery swap stations might be needed for very long belts.
• Sensor resolution – to accurately identify a single bad roller, the drone’s camera must have sufficient resolution or the drone must fly very close to each idler, which can be tricky in terms of flight path.
• Regulatory and safety concerns – especially in underground mines or around personnel, drone use may be restricted.

As of now (2025), drones for conveyor idler monitoring are mostly in R&D or pilot phase, and they tend to focus on thermal/visual detection (later-stage fault identification). They are excellent for periodic inspections over large areas but are not continuous monitors.

An important emerging solution that complements many of the above is the use of artificial intelligence (AI) and computer vision to analyse conveyor idler data. This is especially applied to visual and thermal imaging. Researchers have developed deep learning models (e.g. convolutional neural networks) to automatically detect a faulty idler from images of the conveyor. One recent study used an improved YOLOv5 neural network to perform real-time detection of idler faults from camera footage. Such a system would use a video camera overlooking the belt and idlers and the AI model would flag anomalies – for instance, detecting if an idler is not rotating when it should, has an unusual wobble, or if there is visible damage. Another approach is thermal image analysis: using infrared images, AI can distinguish a true overheated idler from other heat sources or reflections. For example, Siami et al. (2022) developed a CNN-based method to identify overheated idlers in complex thermal backgrounds, achieving robust detection even when sun-heated ground or equipment caused confusing thermal patterns [27]. They and others have also explored image fusion (combining RGB visual images with IR thermal images) to improve fault identification accuracy. Data analytics capability is a key differentiator among idler monitoring technologies: simpler systems might only send an alarm if a threshold is exceeded, whereas advanced systems use machine learning and data fusion to predict failures with fewer false alarms. For instance, the Smart-Idler’s analytics fuse vibration and temperature data to diagnose failure modes and trend the condition over time [2], [3], [12], [5], and its cloud platform can notify maintenance of rollers that are deteriorating long before they fail. In contrast, a basic DTS fibre or a temperature sensor network might just give a yes/no alarm for high temperature. Going forward, AI-driven analytics (either at the edge, on devices, or in the cloud) are increasingly integrated with these systems to improve detection accuracy [27]. Many vendors now offer software that performs trend analysis, compares each idler’s readings against baselines or peer units, and even estimates remaining life. This data analytics capability is a crucial part of CBM solutions and often determines how useful the raw sensor data ultimately is to mine operators.

Condition-based monitoring of conveyor idlers has advanced significantly, from basic temperature switches to sophisticated IoT rollers and fibre-optic sensing. Commercial solutions like wireless sensor networks and smart idlers are already improving uptime and safety in real operations [5], [18], [20], [21], [26]. Meanwhile, experimental and emerging methods (robotics, AI vision) promise to further reduce manual effort and catch failures with even greater reliability. Modern idler CBM systems leverage cloud platforms, machine learning, and even digital twins to predict failures and schedule replacements optimally [1], [12], [27]. In summary, conveyor idler CBM has evolved into a multi-faceted domain – from smart rollers “reporting” their health, to fibre-optic nerves along the belt, to robotic caretakers – all aimed at ensuring the conveyor systems stay healthy and prevent small idler issues from growing into big problems.

For a detailed comparison see part two of this research: “Comparative Analysis of Idler Monitoring Systems”.

[1] M. J. Ibarra-Cabrera et al., “An IoT Solution to Detect Overheated Idler Rollers in Belt Conveyors,” IJACSA, 2024.

[2] Vayeron Pty Ltd., “Smart-Idler® Failure Detection Efficacy Analysis,” 2020.

[3] Delft University of Technology & Rulmeca Group, “Investigation of Smart Conveyor Idler Monitoring Technologies,” 2017.

[4] Liu, X.; Pang, Y.; Lodewijks, G.; He, D. Experimental research on condition monitoring of belt conveyor idlers. 2020.

[5] Anibal S. Morales, Pablo Aqueveque, Jorge A. Henriquez “Technology Review of Idler Condition Based Monitoring Systems,” 2017.

[6] Scott Automation, “Robotic Idler Predict System Whitepaper,” 2023.

[7] Micomo Pty Ltd., “Roller Condition Monitoring (RCM) System Description and Case Studies,” 2014.

[8] P. Dabek et al., “Automatic Detection of Overheated Idlers Using IR and RGB Imaging,” 2022.

[9] Yokogawa Electric Corp., “Fiber Optic DTS for Conveyor Fire Prevention,” 2013.

[10] Artur Küpper GmbH & Co. KG, “Sensor-Integrated Idler Roller Development,” 2019.

[11] Mining3, “Conveyor belt monitoring for wear detection,” 2023.

[12] Smartidler.com, “Technical specifications and product pages.”

[13] Vayeron Pty Ltd, “An idler, a method for monitoring a plurality of idlers, and a conveyor system”, AU2014/050246, Australia, 2015

[14] Vayeron Pty Ltd., “Smart Idler – Effect on Project Economics,” Vayeron, 2025.

[15] Hawkfiber.com / Hawk Measurement, “Fiber-based conveyor sensing.”

[16] Hawk Measurement Systems, “Conveyor Fire Prevention via Thermal Sensing,” 2020.

[17] Australian Government, “Underground Conveyor Safety Standards AS 1755,” 2020.

[18] Rio Tinto, “Smart Idler Rollout and Maintenance Savings,” Case Study, 2018.

[19] J. Widodo and B. Yang, “A review of vibration and acoustic measurement methods for monitoring rolling element bearing condition,” Mechanical Systems and Signal Processing, July 2011.

[20] Glencore, “Underground Coal Mine,” Case Study, 2019.

[21] Paul Moore, Head to Tail, International Mining, February, p.46-63, 2024.

[22] Sandpit Innovation Pty Ltd, Lewis Australia Pty Ltd, “Conveyor Belt Roller Replacement”, WO2013138841, Australia, 2013.

[23] K.R. Thieme, Report: “Economic Justification of Automated Idler Roll Maintenance Applications in Large-Scale Belt Conveyor Systems”, Delft University of Technology, Netherlands, 2014.

[24] Intium Energy Limited, “Vibration detection system, apparatus and method”, WO 2012122597 A1 | PCT/AU2012/000263, Australia, 2012.

[25] Ben Yang Yang, “Fibre Optic Conveyor Monitoring System”, MSc degree thesis, The University of Queensland, Australia, 2014.

[26] Industrial Technologies Supply, itecsu.com, “Remote conveyor rollers monitoring system – Smart-Idler, Interview with Mark Walter at IMARC”, 2023

[27] Siami-Namini, S., et al., ‘Deep learning for conveyor idler fault detection using YOLOv5,’ Conference, 2022.

[28] M. Kang, S. Lee, and J. Kim, “Acoustic-based fault diagnosis for rotating machinery in noisy environments using cyclostationary analysis,” Mechanical Systems and Signal Processing, 2020.

[29] A. Morales, S. Seshadri, and J. Brayan, “Acoustic emission-based condition monitoring of conveyor idler rollers using mobile robots,” IEEE Transactions on Industry Applications, pp. 2639–2647, May–June 2017.

[30] SKF Group, “SKF Idler Sound Monitor Kit: Field evaluation of conveyor idler condition using ultrasonic listening technology,” SKF Mining Industry Solutions, 2019.