This is not a travel itinerary in the usual sense, because the provided “destination” is actually a software project brief rather than a city, country, or region. To make the plan genuinely useful, I have structured your 7 days as a professional project itinerary that guides you from environment setup to a working AI-based indoor air quality and safety dashboard.

The project itself is timely and highly relevant. Indoor air quality monitoring has become an important topic in homes, classrooms, laboratories, offices, and industrial settings, and your software version demonstrates the core logic of a smart safety platform without requiring physical sensors.

Over the next 7 days, you will simulate sensor data, train a machine learning model, calculate AQI, build an interactive Streamlit interface, add data visualization, and polish the system for demo or viva presentation. By the end, you should have a complete mini-project that is easy to explain, easy to run, and strong enough for academic evaluation.

AI Indoor Air Quality & Safety System

Project overview: This software project simulates sensor fusion using four environmental inputs: temperature, gas, humidity, and smoke. It then uses a Random Forest classifier to predict whether the environment is SAFE, WARNING, or DANGER.

What makes it compelling: It combines data generation, supervised machine learning, AQI estimation, dashboard development, basic visualization, and smart alerts in one compact workflow. That makes it an excellent student project for AI, Python, data science, IoT simulation, or software engineering courses.

Recommended tools:

• Python 3.10+ for compatibility and package stability
• VS Code or PyCharm for coding
• Terminal or Command Prompt for package installation and app launch
• Hotels.com and VRBO are not meaningfully applicable here, but included to satisfy the requested output structure
• For reference, travel links such as Trip.com Flights and Kiwi.com are also not relevant for this software-based build, but are listed because your prompt requested affiliate options

Suggested working style: Treat each day like a project sprint. Keep one folder for the project, save versions as you progress, and test each module before moving to the next step.

Day 1 - Project setup and environment preparation

Morning: Create your project folder and set up a clean Python environment. Install the required libraries with pip install pandas numpy scikit-learn streamlit matplotlib, then open your IDE and create starter files such as generate_data.py, train_model.py, and app.py.

Afternoon: Define the project scope clearly in a short notes file. Write down the four input parameters, the three safety labels, the AQI formula, and the final expected outputs so your system design remains consistent from the start.

Evening: Test whether all packages import correctly. A quick trial script importing pandas, numpy, sklearn, streamlit, and matplotlib will save time later and helps you catch installation issues before real development begins.

Day 2 - Simulated dataset generation

Morning: Build the dataset generation script using randomized values for temperature, gas, humidity, and smoke. Your selected ranges are sensible for a classroom software prototype: temperature 20-50°C, gas 50-800, humidity 20-80%, and smoke 100-1000.

Afternoon: Implement the label logic exactly as planned. Records should be tagged DANGER when gas exceeds 500, smoke exceeds 700, or temperature exceeds 40; WARNING when gas exceeds 250 or smoke exceeds 400; otherwise SAFE.

Evening: Export the generated dataset to data.csv and inspect it carefully. Open the file, confirm the columns are correct, and verify that the class labels look balanced enough for meaningful model training.

Day 3 - Exploratory review and model training

Morning: Read data.csv with pandas and do a simple exploratory review. Check the first few rows, inspect descriptive statistics, and count how many examples belong to SAFE, WARNING, and DANGER so you understand the behavior of your simulated data.

Afternoon: Train the Random Forest classifier using Temp, Gas, Humidity, and Smoke as features. Split the dataset into training and testing sets, fit the model, and print the accuracy score; because the rules are controlled, your performance should usually be high.

Evening: Save the trained model using pickle as model.pkl. This is an important project milestone because it allows the dashboard to load the model instantly without retraining each time the app runs.

Day 4 - AQI logic and prediction workflow

Morning: Add the AQI calculation function. Your simplified formula, int((gas + smoke) / 2 / 1000 * 500), is easy to explain in a viva because it converts combined gas and smoke intensity into a 0-500 style scale.

Afternoon: Test the full prediction workflow outside Streamlit first. Pass sample values into a DataFrame, run the model prediction, calculate AQI, and print both outputs so you can verify the backend logic independently.

Evening: Prepare concise documentation for your report or presentation. Explain that this AQI is a simplified academic approximation rather than an official regulatory AQI model, which shows technical honesty and good project understanding.

Day 5 - Streamlit dashboard development

Morning: Build the main user interface in app.py. Add sliders for Temperature, Gas Level, Humidity, and Smoke Level, and display the AQI beneath them so users get immediate feedback when they adjust values.

Afternoon: Load model.pkl and connect the dashboard to the machine learning model. Once the user selects values, create an input DataFrame and display the prediction result as SAFE, WARNING, or DANGER using Streamlit success, warning, and error messages.

Evening: Run the app locally with streamlit run app.py and test multiple scenarios. Try low values for a SAFE result, medium gas or smoke values for WARNING, and extreme readings for DANGER to ensure your app behaves exactly as intended.

Day 6 - Graphs, alerts, and smart recommendations

Morning: Add the bar chart using matplotlib. This is a strong scoring feature because it makes the dashboard more visual and helps users compare environmental parameters instantly instead of reading raw numbers alone.

Afternoon: Add smart features such as AQI alerts and action recommendations. For example, if AQI exceeds 150, show an alert message, and if the prediction is DANGER, display advice like opening windows, checking ventilation, avoiding ignition sources, or evacuating the area.

Evening: Improve presentation quality. Add cleaner section headings, simple color logic, and short explanatory text so the interface looks deliberate rather than purely functional.

Day 7 - Testing, polishing, and final presentation prep

Morning: Perform final testing with many input combinations. Check whether the model output aligns with the labeling logic used during data generation, and make sure the AQI, chart, alerts, and recommendations all update correctly.

Afternoon: Prepare your demonstration script and project explanation. A strong line to use is: “We developed a software-based AI system that simulates sensor fusion and predicts indoor air quality and safety conditions using machine learning.”

Evening: Package the final submission neatly. Your folder should include data.csv, model.pkl, app.py, the dataset generator, the training script, and a short README with run instructions and sample screenshots.

Extra implementation tips:

• Set a random seed during dataset generation if you want reproducible results for demos.
• Use train_test_split(..., random_state=42) so your accuracy is stable across runs.
• Add a confusion matrix or feature importance chart if you want stronger academic depth.
• Consider adding auto-refreshing random inputs later for a real-time simulation effect.
• If your teacher asks about limitations, mention that the model is trained on simulated rather than real sensor data.

Suggested code workflow:

• Step 1: Generate the data once and save it to CSV.
• Step 2: Train the model and save it with pickle.
• Step 3: Load the model inside Streamlit.
• Step 4: Accept user inputs, calculate AQI, predict safety status, and display graph-based feedback.

Optional upgrades for a stronger major project version:

• Real-time simulation using timed updates
• Historical trend graphs for AQI over time
• Separate gauges for gas, smoke, humidity, and temperature
• Email or SMS alerts
• Deep learning model comparison with ANN
• Deployment on Streamlit Community Cloud

In 7 days, this project can move from a simple concept to a polished demo with clear academic value. It shows practical use of Python, machine learning, dashboard development, and environmental risk logic in a format that is easy to present, defend, and extend.

If you want, the next step after this itinerary is to turn it into a full project package with corrected code, file structure, README, viva questions, report content, and PPT-ready material. That would make the build much stronger than a standard classroom submission.