Challenge 0: Environment Setup

Welcome to your very first challenge! In this challenge, we’ll set the foundation for the hackathon. You’ll deploy the required Azure resources, set up your development environment, and prepare the assets needed for the next challenges. By the end, you’ll have everything ready for the rest of the hackathon.

If something isn’t working as expected, please let your coach know.

Expected duration: 45-60 min

Prerequisites:

Azure Subscription with permissions to create resources
GitHub Account to fork the repository
GitHub Codespaces access
Azure CLI (pre-installed in Codespaces)

🎯 Objective

The goals for this challenge are:

Provision the Azure resources needed for the upcoming challenges.
Seed sample data for the tire factory predictive maintenance multi-agent system.

🧭 Context and Background

A number of different Azure Resources are used as part of this hackathon.

The following image illustrates the setup:

All resources reside in a single resource group.

A Foundry Account with a Foundry Project with model deployments for gpt-4.1, gpt-4o-mini and text-embedding-ada-002
API Management with two API proxies that read data from Cosmos DB
Cosmos DB with a database including data for machines, technicians, work orders, and more
Storage Account with knowledge base wiki articles in Markdown
Application Insights and Log Analytics for logging
AI Search to query wiki data
Container Registry and Container Apps Environment for running containers

Sample data is seeded into different Cosmos DB containers that will be queried during the exercises.

Cosmos DB Data Model (7 Containers)

Container	Partition Key	Purpose	Sample Count
`Machines`	`/type`	Equipment definitions	5 machines
`Thresholds`	`/machineType`	Operating limits	13 thresholds
`Telemetry`	`/machineId`	Sensor readings	10 samples
`KnowledgeBase`	`/machineType`	Troubleshooting	10 articles
`PartsInventory`	`/category`	Spare parts	16 parts
`Technicians`	`/department`	Maintenance staff	6 technicians
`WorkOrders`	`/status`	Maintenance history	5 work orders

Sample Machines (5 Production Units)

Tire Curing Press A1 (machine-001)
- Type: tire_curing_press
- Status: Operational
- Operating Hours: 12,450
- Cycles Completed: 45,680
- Key Metrics: Temperature (165–175°C), Pressure (150–190 bar)
Tire Building Machine B1 (machine-002)
- Type: tire_building_machine
- Status: Operational
- Tires Built: 67,840
- Key Metrics: Drum vibration, Ply tension
Tire Extruder C1 (machine-003)
- Type: tire_extruder
- Status: Operational
- Total Output: 1,245 tons
- Key Metrics: Barrel temperature, Extrusion pressure
Tire Uniformity Machine D1 (machine-004)
- Type: tire_uniformity_machine
- Status: ⚠️ Maintenance Required
- Tires Inspected: 98,450
- Key Metrics: Force variation, Balance
Banbury Mixer E1 (machine-005)
- Type: banbury_mixer
- Status: Operational
- Batches Completed: 15,670
- Key Metrics: Mixing temperature, Power consumption

Telemetry Samples (with Anomalies)

The seeded data includes warning conditions to test your agents:

🔴 Machine 001: Temperature 179.2°C (⚠️ exceeds 178°C warning)
🔴 Machine 002: Drum vibration 3.2 mm/s (⚠️ exceeds 3.0 mm/s)
🔴 Machine 003: Throughput 640 kg/h (⚠️ below 650 kg/h minimum)
🔴 Machine 004: Radial force variation 105N (⚠️ exceeds 100N)
🔴 Machine 005: Multiple warnings (temp, power, vibration)

Knowledge Base (10 Troubleshooting Sample Guides)

Sample articles include:

Curing temperature excessive
Building drum vibration
Extruder barrel overheating
High radial force variation
Mixer vibration issues

Each article contains:

Symptoms & possible causes
Diagnostic steps
Solutions & repair procedures
Estimated repair times

Parts Inventory (16 sample Spare Parts)

Categories include:

Bladders, seals, and heating elements
Bearings and servo motors
Sensors and load cells
Extruder screws and dies
Mixer rotor tips

Sample parts with low stock trigger reorder alerts.

Technicians (6 Specialists)

John Smith - Senior Tire Equipment Technician
Maria Garcia - Building Machine Specialist
David Lee - Quality Systems Technician (⚠️ on assignment)
Sarah Johnson - Electrical Technician
Michael Chen - Mixing & Extrusion Technician
Jennifer Rodriguez - Mechanical Technician

✅ Tasks

Task 1: Fork the repository

Before you start, please fork this repository to your GitHub account by clicking the Fork button in the upper right corner of the repository's main screen (or follow the documentation). This will allow you to make changes to the repository and save your progress.

Task 2: Launch the development environment

GitHub Codespaces is a cloud-based development environment that allows you to code from anywhere. It provides a fully configured environment that can be launched directly from any GitHub repository, saving you from lengthy setup times. You can access Codespaces from your browser, Visual Studio Code, or the GitHub CLI, making it easy to work from virtually any device.

To open GitHub Codespaces, click on the button below:

Please select your forked repository from the dropdown and, if necessary, adjust other settings of GitHub Codespace.

Note

If GitHub Codespaces is not enabled in your organization, see enabling or disabling Codespaces for your organization. If you can’t change your organization’s settings, you can also create a free personal GitHub account. The GitHub Free plan includes 120 core hours per month, equivalent to 60 hours on a 2-core machine, along with 15 GB of storage.

⏱️ Initial creation of the Codespace takes approximately 10 minutes

Tip

While you are waiting for the Codespace creation to complete, take some time to familiarize yourself with the hackathon scenario and architecture.

Task 3: Log in to Azure

Once your Codespace has started, log in to the Azure CLI by running the command below in the terminal and following the prompts.

az login --use-device-code

Task 4: Deploy Resources

In this step, you create the resources that will be used throughout the day.

Important

Depending on the setup for the hackathon, the Azure resources might already have been provisioned for you and you can then skip this step. Check with your hackathon coach what is applicable for you.

Deploy Azure resources

# Ensure Microsoft.AlertsManagement resource provider is registered for use in the subscription
az provider register --namespace Microsoft.AlertsManagement

# Ensure you are located in the challenge-0 directory
cd challenge-0

# Make resource group name easy to identify. Use your initials or other identifier (e.g., "jd" for John Doe)
export RG_SUFFIX="<initials>"

# Set variables with your initials as suffix
export RESOURCE_GROUP="rg-tire-factory-hack-${RG_SUFFIX}"
export LOCATION="swedencentral"

# Create resource group
az group create --name $RESOURCE_GROUP --location $LOCATION

# Deploy infrastructure
az deployment group create \
  --resource-group $RESOURCE_GROUP \
  --template-file infra/azuredeploy.json \
  --parameters location=$LOCATION

⏱️Deployment takes approximately 5-10 minutes.

Task 5: Verify the creation of your resources

Go to the Azure Portal and find your resource group, which should now contain resources like this:

Task 6: Retrieve keys for environment variables

After deploying resources, configure environment variables in the .env file. Ensure you're logged into Azure CLI, then run the get-keys.sh script to automatically populate the required values.

Important

Wait until all Azure resources are successfully deployed before starting this task. Otherwise, the environment variables may not be extracted correctly. If the environment is pre-created for you, the resource group name will be provided by the hackathon coaches. Set it using export RESOURCE_GROUP='your predefined resource group name'

# Ensure you are in the challenge-0 directory
cd challenge-0

# Extract connection keys
./get-keys.sh --resource-group $RESOURCE_GROUP

# Verify .env file. No entries should be empty
cat ../.env

# Make environment variables available in the shell
export $(cat ../.env | xargs)

Tip

Keep your .env file handy throughout the hackathon. You need to re-export the environment variables each time you open a new shell or when you resume a stopped codespace.

Caution

For convenience, we use key-based authentication and public network access to resources in this hackathon. In real-world implementations, you should consider stronger authentication mechanisms and additional network security. Never commit the .env file to the repository. This repo already includes .env in .gitignore, but if you rename the file you may need to add the new name to .gitignore as well.

Task 7: Seed Factory Sample Data

As mentioned in Context and Background, there are several data sources used throughout the hackathon. Run the script below to upload data to Cosmos DB and the Storage Account, and to create the required APIs in API Management.

# Run data seeding script
./seed-data.sh

Task 8: Assign additional permissions

To perform certain tasks in the hackathon, you need the following permissions:

Azure AI Developer on the Foundry project resource (agent/project operations)
Azure AI User on the AI Services resource (agent create/run operations)
Cognitive Services OpenAI Contributor on the Azure OpenAI resource (calling chat completions)

Important

Depending on the setup for the hackathon, the Azure roles might already have been assigned to you in advance, and you can then skip this step. Check with your hackathon coach what is applicable for you.

Assign permissions

# Get your Entra ID (AAD) user object ID
ME_OBJECT_ID="$(az ad signed-in-user show --query id -o tsv)"

# Assign "Azure AI Developer" at the AI Foundry Project resource scope
az role assignment create \
  --assignee-object-id "$ME_OBJECT_ID" \
  --assignee-principal-type User \
  --role "Azure AI Developer" \
  --scope "$AZURE_AI_PROJECT_RESOURCE_ID"

# Assign "Azure AI User" at the AI Services resource scope
# This is required for agent operations (create/run agents)
# Derive the AI Services ID from the project resource ID
AI_SERVICES_ID="${AZURE_AI_PROJECT_RESOURCE_ID%/projects/*}"

az role assignment create \
  --assignee-object-id "$ME_OBJECT_ID" \
  --assignee-principal-type User \
  --role "Azure AI User" \
  --scope "$AI_SERVICES_ID"

# Assign "Cognitive Services OpenAI Contributor" at the Azure OpenAI resource scope
# This is required for data-plane calls like: /openai/deployments/{deployment}/chat/completions
OPENAI_RESOURCE_ID="$(az cognitiveservices account show --name "$AZURE_OPENAI_SERVICE_NAME" --resource-group "$RESOURCE_GROUP" --query id -o tsv)"

az role assignment create \
  --assignee-object-id "$ME_OBJECT_ID" \
  --assignee-principal-type User \
  --role "Cognitive Services OpenAI Contributor" \
  --scope "$OPENAI_RESOURCE_ID"

# Refresh your credentials with the new permissions
az login --use-device-code

Role assignments can take 5–10 minutes to fully propagate. If you still see PermissionDenied errors after assigning roles, wait a few minutes, then run az login --use-device-code again and re-export your environment variables.

🎉 Congratulations! Your sample tire factory environment is ready.

🚀 Go Further

Note

Finished early? These tasks are optional extras for exploration. Feel free to move on to the next challenge — you can always come back later!

Run Sample Queries

If you want to verify or explore the seeded data, here are some sample queries you can run against the Cosmos DB. This can be done via the Azure Portal Data Explorer. As shown below:

Find machines with warnings in the Telemetry container

SELECT c.machineId, c.status, c.alerts FROM c WHERE c.status = "warning"

Get curing press thresholds in the Thresholds container

SELECT c.metric, c.normalRange, c.warningThreshold, c.criticalThreshold
FROM c
WHERE c.machineType = "tire_curing_press"

Find available technicians with curing press skills in the Technicians container

SELECT c.name, c.skills, c.available
FROM c
WHERE ARRAY_CONTAINS(c.skills, "tire_curing_press") 
  AND c.available = true

🛠️ Troubleshooting and FAQ

Problem: ARM template deployment fails

# Check deployment errors
az deployment group show \
  --resource-group $RESOURCE_GROUP \
  --name azuredeploy \
  --query properties.error

# Register missing providers (if needed)
az provider register --namespace Microsoft.AlertsManagement
az provider register --namespace Microsoft.App

Problem: seed-data.sh script fails

# Verify Cosmos DB is ready
az cosmosdb show \
  --name $COSMOS_NAME \
  --resource-group $RESOURCE_GROUP \
  --query provisioningState

# Check if containers exist
az cosmosdb sql container list \
  --account-name $COSMOS_NAME \
  --resource-group $RESOURCE_GROUP \
  --database-name "FactoryOpsDB"

# Re-run seed script (idempotent)
bash challenge-0/seed-data.sh

Problem: Permission denied on seed-data.sh

# Add execute permission to the script
chmod +x challenge-0/seed-data.sh

🧠 Conclusion and Reflection

This forms the foundation for your multi-agent predictive maintenance hackathon system.

Time to build some intelligent agents!

Important

This hackathon uses simplified authentication for learning purposes. Production systems should implement:

Managed identities instead of keys
Private endpoints for network security
Azure Key Vault for secrets management
RBAC for fine-grained access control

If you want to learn more about what you covered in this challenge, check out the links below:

Next step: Challenge 1 - Building Agent Framework Agents for Anomaly Classification and Fault Diagnosis

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Challenge 0: Environment Setup

🎯 Objective

🧭 Context and Background

✅ Tasks

Task 1: Fork the repository

Task 2: Launch the development environment

Task 3: Log in to Azure

Task 4: Deploy Resources

Task 5: Verify the creation of your resources

Task 6: Retrieve keys for environment variables

Task 7: Seed Factory Sample Data

Task 8: Assign additional permissions

🚀 Go Further

Run Sample Queries

🛠️ Troubleshooting and FAQ

🧠 Conclusion and Reflection

FilesExpand file tree

README.md

Latest commit

History

README.md

File metadata and controls

Challenge 0: Environment Setup

🎯 Objective

🧭 Context and Background

✅ Tasks

Task 1: Fork the repository

Task 2: Launch the development environment

Task 3: Log in to Azure

Task 4: Deploy Resources

Task 5: Verify the creation of your resources

Task 6: Retrieve keys for environment variables

Task 7: Seed Factory Sample Data

Task 8: Assign additional permissions

🚀 Go Further

Run Sample Queries

🛠️ Troubleshooting and FAQ

🧠 Conclusion and Reflection