To achieve the goal of identifying undiscovered huacas in Lima using artificial intelligence, the research will follow a structured methodology that combines data collection, model training, validation, and field verification. The key steps are as follows:

Identification of Existing Huacas

• Use online mapping tools such as Google Maps and Apple Maps, along with historical and archaeological data, to identify and mark the locations of currently known huacas in Lima
• xpand this process by referencing similar archaeological structures in other regions of the world to enrich the dataset and improve model generalization

Dataset Creation and Annotation

• Collect satellite imagery and aerial photographs of the identified huacas and label these images to create a dataset. This step will serve as the foundation for training the AI model
• Use high-resolution satellite services such as Google Earth, Sentinel Hub, or Maxar to obtain consistent and high-quality imagery
• Peru has a sattelite for scientific and military purposes. Approach the government to get this data

Model Training

• Train a machine learning model to recognize patterns and features specific to huacas
• The project will utilize supervised learning, where labeled data (images of huacas and non-huacas) are used to teach the model to differentiate between these categories
• For the model framework, consider using PyTorch or TensorFlow. Both are widely supported for image recognition tasks. PyTorch is particularly recommended for its flexibility and ease of debugging during research-oriented projects

Inference and Predictions

• Apply the trained model to online maps and satellite images of Lima to predict and identify locations where undiscovered huacas might exist
• Leverage online GIS platforms (Geographic Information Systems) to integrate satellite imagery with prediction data, aiding visualization and analysis

Field Verification

• Form a field team to visit the predicted sites and verify whether the identified structures are indeed huacas. This step is crucial for validating the accuracy of the model and refining its predictions

Iterative Model Improvement

• Based on the results from field verification, retrain and improve the model. Incorporate new data from verified huacas to enhance its accuracy and reliability
• Implement techniques like transfer learning or hybrid models (e.g., combining supervised learning with unsupervised methods like clustering) to optimize performance

The outcomes of this research carry significant value for Peru, as the creation of a digital database of huacas represents a pivotal step toward preserving and promoting the nation's cultural heritage. This data can serve as a critical resource not only for archaeologists and historians but also for scientific communities and private organizations interested in cultural preservation, tourism development, or urban planning.

A particularly noteworthy result would be the identification of a previously undiscovered huaca, which would immediately mark the success of this project. Such a discovery would elevate Peru’s status as a hub for cultural and archaeological innovation, showcasing the potential of AI and machine learning in heritage research.

The results of this study will be shared with the Peruvian government and the Ministry of Culture, reinforcing the importance of integrating advanced technology into national heritage efforts. This collaboration would also contribute to the growing recognition of Peruvian science on a global scale, giving additional value to this pioneering work.

The implications of this research extend far beyond the academic realm, offering opportunities to promote tourism and local economic growth. Among the practical applications are:

Creating a "Ruta de las Huacas"

A dedicated tourist route connecting the most significant huacas, designed to offer an immersive cultural experience.

• Features along the route could include
• Restaurants serving traditional Peruvian cuisine
• Artisan markets showcasing local crafts and products
• Educational centers or museums detailing the history of each huaca
• Recreational areas or guided tours integrating augmented reality (AR) for visitors

Boosting Local and Regional Tourism

By increasing awareness and accessibility, the project has the potential to attract both local and international tourists, leading to increased economic activity. This would translate directly into greater revenue for local communities, businesses, and the tourism sector.

The execution of this project depends on the scope and complexity. Two potential timelines are outlined below:

Fast-Track Implementation (fast implementation)

• Timeframe: 6-8 months
• Focus: Basic dataset creation, initial AI model, and satellite inference for a smaller region of Lima
• Estimated Cost: $50,000–$75,000 (including data acquisition, model training, and field validation)

Comprehensive Implementation (slow implementation)

• Timeframe: 12-18 months
• Focus: Full dataset covering all of Lima, a highly optimized model, extensive field verification, and development of a "Ruta de las Huacas"
• Estimated Cost: $150,000–$200,000 (including high-resolution imagery, advanced AI/ML infrastructure, field validation teams, and integration with tourism plans)

This research highlights the transformative potential of combining AI/ML technology with cultural preservation. By identifying new huacas and valorizing existing ones, the project not only enriches Peru’s historical narrative but also drives economic growth through tourism and community development. With proper support and collaboration, this initiative could position Peru as a global leader in technological applications for heritage, bridging the past and the future for the benefit of all.