The rapid advancement of autonomous vehicles necessitates the development of sophisticated control systems reliant on continuous feedback from tires. Smart tires, integrating strain sensing with traditional tire functionality, offer a solution for continuous monitoring of dynamic parameters such as tire pressure, road friction, and tire deformation. Existing solutions often suffer from limitations including rigid sensors, external power requirements, and complex, costly multi-step fabrication processes. This research tackles these challenges by introducing a novel smart tire system that incorporates several key innovations: 1. **Direct Mask-less 3D Printed Strain Gauges:** Utilizing a graphene-based ink, strain sensors are directly 3D printed, simplifying fabrication and reducing costs. The estimated cost of a single 3D printed sensor is approximately $2.7. 2. **Flexible Piezoelectric Energy Harvester:** A flexible piezoelectric energy harvester scavenges mechanical energy from tire deformation to power the sensing system, eliminating the need for batteries and enhancing long-term reliability. 3. **Secure Wireless Data Transfer:** A secure wireless data transfer system ensures reliable and tamper-proof communication of sensor data to the vehicle's control unit. 4. **Machine Learning for Predictive Data Analysis:** Machine learning algorithms process the sensor data to provide predictive insights into tire behavior and road conditions. This integrated approach significantly advances the design and fabrication of cost-effective smart tires, enabling real-time monitoring and control crucial for the safe operation of autonomous vehicles. The high sensitivity and performance of graphene-based sensors, coupled with the advantages of 3D printing, makes this approach particularly promising for large-scale deployment.

Prior research on integrating wireless sensors into tires for measuring dynamic mechanical parameters has primarily focused on rigid sensors requiring external power. These approaches often involve time-consuming multi-step fabrication processes, resulting in increased complexity and cost. The use of graphene-based materials for sensors has shown promise due to their high performance and sensitivity. However, challenges remain in developing cost-effective, self-powered, and easily integrable solutions. This study builds upon these advancements by leveraging 3D printing technology and piezoelectric energy harvesting to overcome limitations in existing systems. Existing Tire Pressure Monitoring Systems (TPMS) provide limited functionality, low transmission rates, and lack security measures, highlighting the need for an advanced system capable of real-time monitoring and secure data transmission.

The research employed a multifaceted approach involving materials synthesis, 3D printing, sensor characterization, field testing, modeling, and machine learning. **1. Ink Preparation and 3D Printing:** A graphene-based ink was developed using graphite powder, sulfuric acid, phosphoric acid, potassium persulfate, and hydrochloric acid. The ink was used in an aerosol deposition (AD) 3D printing process to fabricate the strain sensors on Kapton film. The printing process parameters, such as flow speed and viscosity, were optimized to achieve a homogenous microstructure. After printing, the graphene oxide (GO) sensors were chemically reduced to reduced graphene oxide (rGO) to enhance conductivity. **2. Materials Characterization:** The printed sensors underwent characterization using various techniques. Transmission electron microscopy (TEM) analyzed the morphology of GO sheets. X-ray diffraction (XRD) and Raman spectroscopy characterized the structural differences between GO and rGO, confirming the successful reduction process. Atomic force microscopy (AFM) examined the surface morphology of the 3D-printed rGO. **3. Piezoresistive Sensor Characterization:** The piezoresistive properties of the sensors were evaluated using a lab setup, measuring the change in resistance under tensile and compressive strain. A linear relationship between resistance and strain was observed up to ~0.7% strain. Time-dependent resistance changes were also investigated. **4. Field Testing:** The 3D-printed sensors were integrated onto a tire and tested on a mobile test rig under varying driving speeds, normal loads, and tire pressures. Data acquisition was performed using LABVIEW. The test rig operated between two parking areas, covering a substantial distance to assess the sensor's reliability and durability. Three zones of strain in a tire (two compressive and one tensile) were identified and analyzed. **5. Tire Modeling:** A tire model was developed to simulate tire deformation under various conditions. The model considered the tire as a ring suspended by radial and tangential springs, and the natural frequencies were estimated by solving the equation of motion using the Hamiltonian principle. The simulated strain was compared to experimental results. **6. Machine Learning:** A machine learning algorithm, using a two-layer feed-forward neural network, was developed to estimate tire pressure based on sensor data, tire normal load, and longitudinal velocity. The algorithm's performance was evaluated using training, validation, and testing datasets. **7. Power Management and Secure Data Transfer:** A piezoelectric energy harvester, utilizing a Polyvinylidene Fluoride (PVDF) patch, was integrated to generate power for the wireless data transfer system. A novel energy-efficient security mechanism was implemented to reduce the computational overhead of secure data transmission, exploiting the fact that sensor measurements often remain unchanged for extended periods in normal operating conditions.

The key findings of the study are: 1. **Successful Fabrication of 3D Printed Graphene Sensors:** The researchers successfully developed a method for 3D printing graphene-based strain sensors using an aerosol deposition technique. The sensors exhibited a wrinkled microstructure that allowed them to withstand large deformations without failure, making them suitable for integration into tires. 2. **Reliable Sensor Performance in Real-World Conditions:** Field testing of the sensors on a mobile test rig demonstrated their ability to accurately measure tire-road interactions under varying driving conditions. The sensor output voltage varied consistently with changes in speed, normal load, and tire pressure. The sensors showed remarkable durability, undergoing more than 8000 cycles during the field tests. 3. **Effective Tire Modeling:** A theoretical model successfully simulated the strain development in a moving tire, accurately predicting the relationship between strain and various parameters (speed, load, pressure). The model's predictions correlated well with experimental results. 4. **Accurate Tire Pressure Monitoring using Machine Learning:** The developed machine learning algorithm accurately estimated tire pressure from sensor data, demonstrating the feasibility of using 3D printed sensors for real-time tire pressure monitoring. The algorithm achieved a high correlation coefficient (above 0.86) and minimal error in pressure estimation. 5. **Feasible Self-Powered Wireless Data Transmission:** The integration of a piezoelectric energy harvester provided a reliable method for powering the wireless data transmission system, eliminating the need for batteries. A novel energy-efficient security mechanism improved the energy efficiency of secure data transmission. 6. **Cost-Effectiveness:** The 3D printing technique significantly reduced the cost of sensor fabrication, making the technology more accessible for widespread implementation in smart tires. The estimated cost of a single sensor was only $2.7.

This research addresses the critical need for reliable, cost-effective, and self-powered sensing solutions for smart tires in autonomous vehicles. The successful development and validation of 3D printed graphene-based strain sensors, coupled with a piezoelectric energy harvester and secure wireless data transfer system, represents a significant advance in this field. The integration of machine learning for predictive analysis further enhances the system's capabilities. The findings demonstrate the feasibility of using this technology for real-time tire pressure monitoring and other critical parameters. The low cost of fabrication, coupled with high sensitivity and durability, makes this approach highly scalable and promising for commercial applications. This work contributes to improving the safety and performance of autonomous vehicles by providing a robust and reliable method for monitoring tire-road interactions.

This study successfully demonstrated the feasibility of using 3D printed graphene-based sensors for smart tire applications in autonomous vehicles. The system's self-powered nature, secure wireless communication, and accurate pressure monitoring capabilities represent significant advancements. Future research directions include exploring different graphene ink compositions to optimize sensor performance, enhancing the energy harvesting efficiency of the piezoelectric system, and developing more sophisticated machine learning algorithms to improve prediction accuracy. Integrating additional sensors for comprehensive tire health monitoring is another promising area of future work.

While the study demonstrates the significant potential of the developed system, some limitations exist. The field tests were conducted under specific environmental conditions, and further testing is required to validate the system's performance under a wider range of conditions. The accuracy of the tire model could be further improved by incorporating additional factors influencing tire dynamics. The currently employed machine learning algorithm could be further optimized by incorporating more data and using advanced neural network architectures to enhance the prediction accuracy and robustness. Finally, long-term reliability and durability under extreme conditions need further investigation.