Economic efficiency of implementation of predictive maintenance in the city taxi fleet: comparative analysis of Tesla Model 3 and Škoda Octavia A7

Abstract

This article determines the cost-effectiveness of implementing a predictive maintenance (PdM) system using the example of a city taxi fleet that uses Tesla Model 3 (electric cars) and Škoda Octavia A7 (gasoline cars).

Traditional maintenance (MOT) on a fixed schedule often leads to unnecessary preventive costs and unexpected downtime due to sudden breakdowns. The implementation of PdM systems based on IoT-sensors, telemetry and data analysis using artificial intelligence (AI) allows you to reduce maintenance costs and increase the reliability of equipment. For intensive operation in an urban taxi fleet, the comparison of traditional and predictive models is critical Important.

PdM systems collect continuous telematics data (temperature, pressure, vibrations, battery status). For example, in the Tesla Model 3, data is collected by standard electronics, while for the Skoda Octavia, external OBD-II adapters can be used, go to the cloud, where ML models are trained (on a Tesla Dojo supercomputer or cloud GPUs). The updated diagnostic algorithms are delivered back to the vehicles via OTA updates. This allows you to detect abnormal patterns of node operation, preventing emergency stops and even remotely solving minor problems.

A comparative analysis of service strategies showed:

1. Routine maintenance has an average risk of unplanned downtime and high costs (due to non-critical scheduled repairs).

2. Reactive maintenance is characterized by maximum downtime and high costs for emergency repairs_..

3. Predictive maintenance minimizes unplanned downtime and reduces maintenance costs by 18–25 %.

The implementation of PdM results in significant savings on maintenance, reduced downtime and life extension.