Remote Troubleshooting APP Development for Excavators Based on Android System

Chao Yang, Chenbo Yin, Shoulei Ma, Donghui Cao, Xiaozhong LiuLecture Notes in Mechanical Engineering 2025 · DOI · OpenAlex

Abstract

Fault diagnosis technology is an important means to maintain the normal operation of the machine. Currently, excavator faults occur frequently, which not only affects the progress of the project but also brings potential safety hazards, and the service engineers need to visit the site for inspection and troubleshooting, which has high time cost and condition limitations. Based on this, a remote troubleshooting tool is developed to improve the troubleshooting efficiency. Firstly, the fault diagnostic device is connected to the excavator to realize the mutual conversion of WIFI signal and CAN signal, and the APP developed by Android system is used to send commands, based on the signal comparison method and the case inference method, the one within a certain threshold error is regarded as no fault, and the other way around is based on the source case to guide the fault case, so as to realize the remote one-button inspection. The APP has been tested by the dSPACE simulation system in the laboratory and the on-board test of SY980, with excellent diagnostic effect, and has been put into use.

Introduction

Excavators operate in notoriously harsh environments where mechanical and electronic failures are frequent. Traditionally, when a machine breaks down, a service engineer must travel to the site—often in remote locations—to perform manual inspections. This process is plagued by high time costs, project delays, and safety risks. While modern excavators have onboard controllers, their arithmetic power is often limited, and they lack intuitive interfaces for deep troubleshooting.

This research addresses these challenges by decoupling diagnostics from hardware. By moving the diagnostic "brain" to an Android app, the system can be updated via the cloud (firmware and fault libraries) without needing to flash the excavator's physical controller. This allows for real-time interaction with the machine's Controller Area Network (CAN) bus via a mobile platform.

Methodology

The researchers developed a full-stack solution comprising hardware integration, mobile software, and cloud-based data management. The architecture is divided into four distinct layers: the Excavator, the Troubleshooter (communication module), the Smartphone (application module), and the Cloud Center.

Layered architecture of the remote troubleshooting system.

Hardware Gateway

A "troubleshooting box" serves as the bridge between the machine and the mobile device. It connects to the excavator’s CAN port and performs mutual conversion between CAN signals (the industry standard for vehicle communication) and WIFI signals, allowing a smartphone to "talk" to the machine.

The hardware gateway (troubleshooter box) and its components.

Software and Logic

The app was developed using JavaScript and compiled via HBuilderX for Android, with a Flask-based backend connecting to a cloud SQL fault database. The system employs two primary diagnostic methods:

  • One-Touch Inspection: The app iterates through all controller ports (specifically DO and PWM ports), sending command signals and comparing feedback against predefined thresholds. If the deviation is too high, a fault is flagged.
  • Case-Based Reasoning (CBR): A correlation matrix ($Q = LR$) maps fault phenomena ($M_p$) to failure causes ($M_c$) based on historical data. This mathematical framework accounts for "many-to-many" relationships where one symptom might point to multiple causes.

Data Sources

The system relies on a proprietary Source Case Library containing fault codes and maintenance steps. During development, real-time simulation data was generated using a dSPACE semi-physical simulation system to mimic excavator characteristics in a controlled environment.

Results

The system underwent two rigorous validation phases:

  1. Laboratory Testing: Using the dSPACE system, faults were simulated by unplugging main pump flow valves. The app successfully detected missing feedback signals and correctly identified the induced faults.
  2. On-Board Field Test: Tested on a Sany SY980, a 98-ton excavator. The app traversed 27 PWM ports and 5 DO ports. Solenoid valves showed physical action within 1–2 seconds of command, and the app synchronously displayed correct fault codes on both the smartphone and the cabin display.

Conclusion and Future Work

While effective, the current logic faces challenges with "intersecting" faults where multiple systems fail simultaneously in ways not yet mapped in the correlation matrix. Additionally, manual updates to the fault library can lag behind the release of new excavator models.

Future work will focus on integrating "Big Data" algorithms to transition from reactive troubleshooting to predictive maintenance. By using trend analysis, the goal is to identify failing components before a breakdown occurs and to expand the database for full coverage across a wider range of heavy machinery.

Summary updated

Abstract

Fault diagnosis technology is an important means to maintain the normal operation of the machine. Currently, excavator faults occur frequently, which not only affects the progress of the project but also brings potential safety hazards, and the service engineers need to visit the site for inspection and troubleshooting, which has high time cost and condition limitations. Based on this, a remote troubleshooting tool is developed to improve the troubleshooting efficiency. Firstly, the fault diagnostic device is connected to the excavator to realize the mutual conversion of WIFI signal and CAN signal, and the APP developed by Android system is used to send commands, based on the signal comparison method and the case inference method, the one within a certain threshold error is regarded as no fault, and the other way around is based on the source case to guide the fault case, so as to realize the remote one-button inspection. The APP has been tested by the dSPACE simulation system in the laboratory and the on-board test of SY980, with excellent diagnostic effect, and has been put into use.
BibTeX
@inproceedings{yang2025remote,
  author = {Chao Yang and Chenbo Yin and Shoulei Ma and Donghui Cao and Xiaozhong Liu},
  title = {{Remote Troubleshooting APP Development for Excavators Based on Android System}},
  booktitle = {Lecture Notes in Mechanical Engineering},
  year = {2025},
  doi = {10.1007/978-981-96-5527-4\_8},
}
Figure