The building inspection process is a critical component in maintaining structural integrity and safety in construction and renovation projects. However, this process is fraught with significant challenges that can hinder its effectiveness. Traditional manual inspections are often labor-intensive, time-consuming, and costly. Inspectors are required to climb ladders, navigate scaffolding, and access potentially hazardous areas, which not only increases the risk of accidents but also drives up the overall cost of inspections.

Inspectors must be highly skilled, as they need to identify and assess various issues, from minor cracks to significant structural defects. This reliance on human expertise can lead to inconsistencies in reporting and potentially overlook critical issues, especially in hard-to-reach locations. Moreover, the logistics involved in scheduling inspections, coordinating with multiple stakeholders, and ensuring compliance with safety regulations can complicate the process further.

To address these challenges, we are developing an autonomous device specifically designed for the inspection and repair of wall cracks. This innovative technology has the potential to revolutionize the building inspection and maintenance landscape. The autonomous device is equipped with advanced sensors and imaging technology, allowing it to conduct thorough inspections without the need for human intervention in hazardous locations.

The Figure 2 below is the the current method used by the restoration team from the Cultural Affairs Bureau of Macau to protect the local heritage sites. This method is also the inspiration of our hardware's desgin and the basis on which it is developed upon.

Key features of our device include:

1. Automated Detection: Utilizing high-resolution cameras and AI algorithms, the device can accurately detect and assess the severity of wall cracks, providing real-time data for further analysis.

2. Remote Operation: The device can be controlled remotely, allowing inspectors to monitor conditions and make decisions without being physically present in dangerous areas.

3. Efficient Repair Mechanism: Once cracks are detected, the device can perform minor repairs autonomously, applying suitable materials to seal cracks and prevent further damage. This proactive approach minimizes the need for extensive manual repairs and reduces downtime.

4. Data Collection and Reporting: The device can compile detailed reports of its findings and repairs, creating a comprehensive digital log that is easily accessible for future reference and compliance audits.

5. Enhanced Safety: By reducing the need for inspectors to access hazardous areas, we significantly lower the risk of accidents, making the inspection process safer for all involved.

By implementing this technology, we aim to streamline operations, enhance safety, and ultimately reduce costs associated with building inspections and maintenance. The transition to autonomous inspection and repair processes not only improves efficiency but also ensures a higher standard of safety and reliability in structural assessments.

This innovation represents a significant step forward in the construction and maintenance industry, paving the way for smarter, safer, and more cost-effective practices. As we continue to develop this technology, we are excited about its potential to transform the way inspections are conducted and how building maintenance is managed.

Our solution focuses on an innovative automatic spraying system specifically designed to efficiently fill cracks in buildings and other structures. This compact system features a robust 3D-printed outer shell, which not only ensures durability but also simplifies the installation process, allowing for quick deployment on-site.

System Overview

The design of the automatic spraying system includes four key components, each playing a vital role in the functionality and effectiveness of the repair process:

• Vibrating Wire Crack Meter:
  This advanced sensor is crucial for accurately monitoring the width of cracks in real-time. By using the vibrating wire technology, it can detect even the slightest changes in crack width, ensuring that any structural movement is promptly identified. This capability allows for timely interventions before minor issues escalate into major problems, ultimately enhancing the longevity of the structure.
• Arduino Main Board:
  Serving as the control core of the system, the Arduino main board processes incoming data from the crack meter and manages the overall operations of the spraying mechanism. With its programmable capabilities, the Arduino can be customized to set specific thresholds for crack width, enabling it to trigger the spraying mechanism only when necessary. This intelligent control reduces material wastage and ensures that repairs are made only when required.
• Precision Spray Nozzle:
  The precision spray nozzle is engineered to dispense microbial repair agents with high accuracy. Designed to adapt to various crack sizes, it ensures that the repair agents are applied evenly and effectively over the detected cracks. The nozzle's design minimizes overspray and maximizes the penetration of repair materials, promoting a strong bond and effective sealing of the cracks.
• Supporting Modules:
  These modules encompass communication and actuation systems that facilitate seamless data transfer and control operations. They enable the Arduino to communicate with the crack meter and the spray nozzle, ensuring synchronized actions. Additionally, these modules can be integrated with wireless communication protocols, allowing for remote monitoring and control of the system from a central management console.

Operation Process

The operation of the automatic spraying system is streamlined to ensure maximum efficiency and effectiveness:

1. Installation: The 3D-printed outer shell is securely attached over the detected crack, creating a stable and controlled environment for the internal components. This design prevents external factors, such as weather conditions, from interfering with the repair process.

2. Continuous Monitoring: The vibrating wire crack meter continuously monitors the crack's width. It sends real-time data to the Arduino main board, which processes the information to determine if the crack width exceeds predefined safety limits.

3. Activation and Repair: When the Arduino detects that the crack has surpassed the designated threshold, it activates the precision spray nozzle. The nozzle then releases the microbial repair agents, filling the crack with a material designed to promote healing and prevent further deterioration. This process minimizes the need for extensive manual repairs and reduces the overall maintenance costs.

Benefits

This automatic spraying system not only enhances the efficiency of structural repairs but also significantly improves safety by minimizing human intervention in potentially hazardous environments. The integration of advanced technology ensures that buildings are maintained at optimal safety levels while reducing the risks associated with traditional repair methods.

By employing this state-of-the-art solution, we are paving the way for smarter building maintenance practices that prioritize both structural integrity and safety.

During our preliminary experiments, we encountered several critical limitations related to our system's functionality that impacted the overall performance and reliability of the autonomous wall crack repair device:

Identified Limitations

• Voltage Detection Issues:
  The multimeter used to measure voltage from the vibrating wire crack meter failed to provide consistent readings. This inconsistency was observed during multiple test runs, where fluctuations in voltage measurements raised concerns about the accuracy and reliability of our data collection process. Reliable voltage readings are crucial for determining the crack width, as they directly influence the system's ability to respond to changes in structural integrity. The erratic nature of the readings suggests potential issues with the sensor calibration, connection stability, or the inherent design of the vibrating wire technology itself.
• Insufficient Voltage Levels:
  The voltage levels recorded by the crack meter were found to be excessively low, which poses a significant challenge. For the Arduino to activate the precision spray nozzle effectively, it requires a minimum voltage threshold to ensure proper operation. The consistently low voltage levels could lead to delayed activation of the nozzle, undermining the intended efficiency of the system. This delay not only affects the timely repair of cracks but also compromises the overall effectiveness of the device in preventing further structural damage. The inability to achieve adequate voltage levels raises questions about the suitability of the vibrating wire crack meter for our specific application.

Implications and Future Directions

These findings indicate a significant limitation in the current design: the vibrating wire crack meter may not be suitable for integration into our autonomous wall crack repair device. To overcome these challenges and enhance the overall effectiveness of our system, we must explore alternative sensor technologies that can deliver reliable and consistent voltage readings.

Potential Alternatives:

• Photoelectric Sensors: These sensors operate on light reflection principles and can provide accurate measurements of crack width without the need for electrical voltage readings. They could offer a more stable alternative, improving response times and reliability.
• Ultrasonic Sensors: By using sound waves to measure distances, ultrasonic sensors could provide precise measurements of crack dimensions, thereby eliminating voltage inconsistencies.
• Camera-Based Systems: Incorporating a camera sensor could enable visual analysis of cracks, allowing for advanced image processing techniques. This approach not only provides accurate detection but also allows for comprehensive data collection, including the ability to assess crack progression over time.

Moving Forward

By addressing these critical challenges, we aim to refine our design and ensure that our autonomous wall crack repair device operates at optimal efficiency. Conducting a thorough evaluation of alternative sensor technologies will be essential in creating a more reliable system. Our goal remains to develop an effective solution that enhances structural safety while minimizing maintenance costs and risks associated with manual inspections.

As we transition to exploring these alternatives, we are committed to testing and validating each option to identify the most suitable technology for our application. This iterative approach will allow us to refine our system and ultimately achieve a robust and reliable autonomous repair solution.

Recognizing the limitations of the crack meter in our hardware, we have decided to replace it with a state-of-the-art camera-based system. This transition aims to enhance the accuracy and reliability of crack detection and repair in our autonomous wall crack repair device.

New Camera-Based System Overview

In this new setup, the camera is strategically positioned to monitor a wall that may have cracks or is at risk of developing them. The camera's field of view serves as a standard reference area. If the visible area of the wall becomes smaller than the camera's total area, the system detects this change, indicating the presence of a crack. The camera captures high-resolution images, and the connected computer, utilizing the processing power of the Raspberry Pi board, analyzes these images to calculate the size and dimensions of the crack. After processing the data, the system intelligently directs the precision spray nozzle to apply microbial agents precisely onto the detected crack.

Components of the New Hardware

1. Camera:

This high-resolution camera connects to the Raspberry Pi 4 board, providing a continuous live feed of the wall. It is capable of capturing detailed images under various lighting conditions, ensuring reliable crack detection. The camera’s field of view can be adjusted to accommodate different wall sizes and configurations, allowing for versatile deployment across various structures.

2. Raspberry Pi 4:

Acting as the control center, the Raspberry Pi 4 powers the camera and nozzles while simultaneously processing the data received from the camera. With its robust processing capabilities, the Raspberry Pi can run image recognition algorithms to detect cracks and calculate their dimensions in real-time. It also manages the communication between different components of the system, ensuring timely responses to detected changes in the wall structure.

3. Precision Spray Nozzle:

This advanced nozzle is engineered to dispense microbial repair agents with remarkable accuracy. By coordinating with the Raspberry Pi, the nozzle can be activated to apply the repair agents directly onto the identified cracks. The design of the nozzle allows for fine-tuned control over the spray pattern and volume, ensuring effective sealing of cracks while minimizing waste of repair materials.

4. Supporting Modules:

These modules include actuation and communication systems that facilitate data transfer and control operations throughout the setup. They enable the Raspberry Pi to relay commands to the spray nozzle, ensuring that repairs are made promptly and efficiently. Additionally, these

modules can incorporate wireless communication capabilities, allowing for remote monitoring and management of the system, enhancing accessibility and control.

Operation Process

The operational workflow of the camera-based system is designed for efficiency:

Monitoring: The camera continuously monitors the wall, capturing images at set intervals or in response to specific triggers (e.g., changes in light or movement).

Image Processing: The Raspberry Pi analyzes the captured images using advanced algorithms to detect cracks. By comparing the visible area of the wall with the camera’s total field of view, it can calculate the presence and size of any cracks that develop.

Repair Activation: Upon detecting a crack, the system processes the dimensions and location, directing the precision spray nozzle to apply the microbial agents directly onto the affected area. This targeted approach ensures effective repairs while reducing the need for manual intervention.

Benefits of the New System

Implementing this camera-based solution enhances the accuracy and reliability of crack detection, allowing for quicker response times and more efficient repairs. By automating the monitoring and repair process, we aim to improve the overall safety and longevity of structures, minimizing the risks associated with traditional inspection methods.

As we move forward with this upgraded system, we are excited about its potential to transform the way we approach building maintenance and structural integrity, making our autonomous wall crack repair device more effective and user-friendly.

Repair-Agent Dosage Prediction

Cracks exhibit a certain width, and for the quantitative replenishment of repair agent, it is necessary to calculate the volume of the cracks. To efficiently perform this calculation, a triangular calculation method is employed. It is assumed that the length of the crack is a fixed value of 1 centimeter, while a normal depth of 0.02 centimeters is adopted based on Macau regulations 60/96/M. Subsequently, the angle of the triangle (as shown in the figure) is measured. By taking the angle as a variable, the volume of the crack can be calculated, thereby enabling the replenishment of the corresponding amount of repair agent. After completing testing, we designed and 3D printed a shell for our hardware, this is shown in the figure 7 below.

Our investigations reveal that the camera-based system outperforms the traditional crack meter in terms of efficiency and reliability. Unlike the crack meter, which relies on a vibrating wire mechanism to gauge crack size, our camera system employs advanced imaging techniques. This innovation effectively addresses one of the major drawbacks of the crack meter: its tendency to produce low voltage readings, which often results in the failure to detect smaller cracks. Such undetected cracks can lead to significant structural issues over time, especially if they remain unmonitored. Furthermore, the inability of the Raspberry Pi 4 board to activate microbial agents in response to these low readings poses an additional challenge. While we initially opted for the crack meter due to its status as an industry standard, our testing revealed that it inadequately activates the nozzles when required. Following discussions with educators at our institution regarding potential alternatives, we confidently transitioned to the camera-based system.

Our new methodology for crack detection utilizes the HSV (Hue, Saturation, Value) color model, which allows for a nuanced analysis of color differences between the cracks and the surrounding wall material. By measuring these differences at the pixel level, we can effectively identify and quantify cracks. The process incorporates binarization techniques to accurately determine the size of each crack. When our system detects that the size of a crack remains static or even increases, it calculates the appropriate amount of repair agent to be applied. This agent is then sprayed onto the crack, ensuring that it shrinks at an optimal pace. In practical applications, the changes in crack size tend to be slow and gradual, making it essential for our algorithm to be finely tuned to enhance the accuracy of these measurements and improve the sensitivity of our crack detection capabilities.

In subsequent discussions with Mr. James Fung, a cultural relic restorer affiliated with the architectural heritage recovery team at the University of Macau and the Imperial Palace cultural heritage inheritance center of Macau, we gained valuable insights into the advantages and limitations of our hardware in real-world applications. He identified several key points:

• 1. Convenience in Application: Our hardware is particularly convenient for use in historic sites like the Ruins of Saint Paul and other tall structures, where traditional methods may be challenging to implement.
• 2. Cultural Sensitivity: It is generally uncommon to directly repair cracks in cultural heritage sites, as such actions could compromise the integrity and historical value of the structures.
• 3. Preference for Lime Water: In the field of cultural heritage restoration, lime water is more widely utilized due to its ability to promote the production of calcium carbonate, which is essential for effective repair.
• 4. Remote Repair Application: Mr. Fung suggested that it would be more advantageous for repair agents to be applied through a remote system rather than automatically. This would allow for greater control and precision during the repair process.
• 5. Semi-Automatic Operation: He also noted that a semi-automatic approach would be preferable, as machinery can malfunction, rendering it unreliable for critical tasks like structural restoration.

In summary, our transition to a camera-based detection system not only enhances our ability to monitor cracks effectively but also aligns with the best practices in the field of cultural heritage preservation. We are committed to refining our algorithm and hardware based on expert feedback to ensure optimal performance in real-world applications.