MODEL FOR ASSESSING THE ROBUSTNESS AND SURVIVAL OF MOBILE ROBOTIC PLATFORMS TAKING INTO ACCOUNT RELIABILITY

Abstract

Background. This publication develops a comprehensive model of reliability, robustness, and survivability of a mobile robotic platform (MRP) with a series-parallel architecture. The aim of the work is to perform mathematical modeling and analysis of the reliability of MRP components based on Weibull and exponential distribution laws, as well as to assess the impact of redundancy and structural redundancy on the overall system resilience to failures. It was established that the critical elements of the MRP are the chassis, motor, and controller, as they are connected in series and have a low level of fault tolerance without redundancy, unlike the sensor and communication components connected in parallel.

Methods. Reliability calculation using series-parallel schemes; evaluation of mean time to failure (MTTF); modeling using Weibull and exponential distributions; sensitivity analysis to evaluate the impact of components on system reliability; assessment of system robustness and survivability under environmental changes or faults.

Results. Based on the constructed models, numerical MTTF values were obtained for all MRP components. The highest fault-free operation times were recorded for the chassis (135,412 hours) and the battery (124,914 hours), while the lowest reliability was demonstrated by the Wi-Fi communication modules (80,000 hours) and BLE modules (70,000 hours). The MTTF for the entire MRP is 21,188 hours, approximately 2.4 years of continuous operation. Sensitivity graphs showed that the chassis, controller, and motor have the greatest impact on overall reliability. These are priorities for redundancy or the creation of degraded modes. The system robustness with one failure (Δ ≤ 1) was evaluated through the decrease in probability of fault-free operation. However, over time (beyond 100,000 hours), reliability drops below 5%, highlighting the importance of accounting for component wear. To increase survivability, the MRP should operate with up to two platform failures, plan routes based on reliability, support degraded mode, and dynamically redistribute tasks.

Conclusions. The proposed model enables highly accurate reliability assessment of the MRP considering its structure, connection types, and functional features of components. The results indicate the need for redundancy of critical components and increased system robustness to ensure failure resilience. The presented methodology can be used for maintenance planning, lifecycle forecasting, and optimization of MRP operation in dynamic environments.