# Predictive Maintenance for Industrial IoT of Vehicle Fleets using Hierarchical Modified Fuzzy Support Vector Machine

**Authors:** Arindam Chaudhuri  
**arXiv:** [1806.09612](https://arxiv.org/abs/1806.09612)  
**Date:** 2018-06-24  
**Summary page:** https://www.summarizepaper.com/en/paper/1806.09612v1/

## Key points

- Connected vehicle fleets are increasingly used in industrial IoT scenarios for machine control and management.
- Predictive maintenance is a key focus to optimize performance, reduce downtime, and minimize labor associated with inspections and preventive maintenance.
- Understanding vehicle fault patterns based on attributes like mileage, age, and type is crucial for effective maintenance strategies.
- The hierarchical modified fuzzy support vector machine (HMFSVM) approach has been proposed to address this challenge.
- HMFSVM model outperforms traditional methods like logistic regression, random forests, and standard support vector machines in predicting and preventing maintenance issues.
- Research conducted at Samsung R & D Institute highlights the effectiveness of HMFSVM through experimental results.
- Leveraging advanced machine learning techniques tailored for industrial IoT applications can enhance predictive maintenance capabilities for optimal fleet performance.

## Summary

The use of connected vehicle fleets is rapidly increasing in various industrial IoT scenarios, where networked smart devices control and manage machines. This trend has opened up new opportunities for predictive maintenance, which aims to optimize performance and reduce downtime and labor associated with inspections and preventive maintenance. Understanding the patterns of vehicle faults based on key attributes such as mileage, age, and type is crucial in this context. To address this challenge, researchers have proposed a hierarchical modified fuzzy support vector machine (HMFSVM) approach. This method offers a unique way to analyze telematics data and improve preventative management strategies for vehicle fleets. The HMFSVM model outperforms traditional methods like logistic regression, random forests, and standard support vector machines in predicting and preventing maintenance issues. A study conducted by Arindam Chaudhuri at Samsung R & D Institute in Delhi, India highlights the effectiveness of the HMFSVM method through several experimental results. By leveraging advanced machine learning techniques tailored specifically for industrial IoT applications, organizations can enhance their predictive maintenance capabilities and ensure optimal fleet performance. This research provides valuable insights into utilizing data-driven solutions for efficient management of connected vehicle fleets in the era of Industry 4.0.