Good first issue: Add one new synthetic manufacturing environment

[programmablemanufacturing]

Goal

Add one new synthetic manufacturing environment to the industrial world model scaffold.

The goal is not to create a validated manufacturing simulator. The goal is to create a simple environment that helps test world-model-style reasoning, uncertainty, planning, and decision-making in manufacturing.

Recommended environment types

Please implement one of the following synthetic environments.

Option 1: Delayed quality environment

A manufacturing process where the true quality result is delayed or expensive to observe.

Example idea:

• the model receives a noisy process signal immediately
• the true quality outcome is revealed after several steps
• measuring true quality has a cost
• the planner must choose actions under uncertainty

Suggested file name:

src/imwm/environments/delayed_quality.py

Option 2: Synthetic process-window environment

A simplified process window where some actions lead to feasible, high-quality outcomes and others lead to higher defect risk.

Example idea:

• action represents simplified process settings
• output includes synthetic quality and defect risk
• there is a hidden or noisy feasible region
• the planner must search efficiently for good actions

Suggested file name:

src/imwm/environments/synthetic_process_window.py

Option 3: Machine degradation environment

A process where the machine or tool condition changes over repeated steps.

Example idea:

• repeated actions increase degradation
• higher degradation increases quality risk
• maintenance or conservative actions can reduce risk
• the planner must balance productivity and quality

Suggested file name:

src/imwm/environments/machine_degradation.py

Minimum behavior

The new environment should include:

• a simple state
• an action input
• a transition rule that updates the state
• a synthetic output such as quality, cost, reward, risk, defect probability, or feasibility
• some noise, uncertainty, hidden state, or delayed feedback

Requirements

• Use only synthetic logic
• Add a short docstring explaining what the environment represents
• Add a small example showing how to instantiate and step through the environment

Example usage

A future user should be able to do something like:

env = DelayedQualityEnvironment()

state = env.reset()

next_state, output = env.step(action)

The exact API can follow the current repo structure.

Done when

• One new synthetic manufacturing environment is added
• The environment can be instantiated
• The environment can be stepped through with an action
• The environment includes uncertainty, noise, hidden state, or delayed feedback
• The code includes a short docstring or comments
• A simple example or smoke test is included

Beginner notes

A simple environment is better than a complicated one.

The most valuable contribution is an environment that makes manufacturing look like a sequential decision problem under uncertainty, not just a static prediction problem.