Factory engineering projects

iolinki helps engineers build real factories faster with AI-assisted IO-Link engineering. Use saved station designs, device data, simulation tests and commissioning exports to prepare real equipment for installation and review.

Open Station Studio and choose Projects to start from an industry template: a conveyor stop, box detection, pump pressure, tank level, or coolant flow. Opening a template and editing a station by hand do not use AI messages.

Sign in with Google to save a project. Studio saves subsequent edits to that project and shows whether the current station is saved. Each changed save creates a revision. Open an earlier revision to recover its station, or save a copy to explore another design.

If another editor saved the project first, Studio keeps your local draft and reports a conflict. Reopen the current project or save your edits as a separate copy. Keep an editable JSON download if you need recovery outside the browser. Saved projects are private to the account that owns them; the existing station share link still encodes the station in its URL fragment.

Build, test and publish the conveyor cell

The Part presence at a stop template has a connected virtual conveyor model. The conveyor advances a part, the configured sensor detects its arrival, and the controller holds and releases the part. Missing detection raises an alarm and stops the motor.

1. Open the template, edit your station and choose Save.
2. Choose Run tests. Studio saves pending edits and tests that exact revision.
3. Review the normal, conveyor-jam and sensor-stuck-off results. A fault test passes when the controller raises the expected arrival-timeout alarm and stops the motor.
4. To share a hosted simulation, acknowledge that anyone with the link can view its station and title, then choose Publish simulation.
5. Open the returned link to run the published station without signing in. The simulation stays pinned to its tested revision when you edit the private project later.

The scenario selector in the live view lets you inspect normal and fault behavior. The results include assumptions and sampled signals: target presence, sensor detection, controller state, motor command and actual motion. You can also download commissioning JSON prepared from the exact tested station.

Publication requires all three tests to pass on the current project revision. Editing the project invalidates the displayed test eligibility; run the tests again before publishing. Other templates remain useful for station design and illustrative Run mode, but this first tested publication model covers a single conveyor stop with an IG6214 presence sensor. Unsupported layouts and invalid settings are reported before tests run.

The model uses a fixed 50ms step and an illustrative 3mm target gap, with material correction and the configured sensing point. It assumes ideal motor/stop actuation and transfer between the placed equipment. It does not model inertia, electrical dynamics, physical IO-Link transport or a production PLC program. Review these assumptions when interpreting results.

Use the same project through MCP

Use the signed-in iolinki MCP connection at https://mcp.iolinki.com/account/mcp, with the same Google account used in Studio. The open endpoint at https://mcp.iolinki.com/mcp can read templates but cannot access an account's projects.

Tool	Use
factory_templates	List the five ready-to-use stations.
factory_project_create	Save a station or an independent template copy.
factory_project_list	List your saved projects.
factory_project_get	Open the current project or a selected revision.
factory_project_save	Save edits using the revision you started from.
factory_project_clone	Copy the current saved station into a new project.
factory_project_test	Run the deterministic conveyor suite on an owned revision.
factory_project_publish	Publish a passing test of the exact current revision as a hosted simulation.

For example:

Read the factory templates and create my own conveyor-stop project. Open it, change the sensor setting, and save using its current revision. Then reopen the previous revision and report what changed.

To publish through MCP, test the saved revision and use the returned test.id with factory_project_publish, supplying the current expected_revision. Ask the assistant to publish only when you intend to make the station and title accessible to anyone with the returned link.

factory_project_save requires expected_revision. A conflict means the project changed after you opened it; inspect the current revision before deciding how to combine the edits. A project copy starts its own history at revision 1. Projects retain up to 60 station components, with a 96KiB station-size limit and 100 projects per account.

Commissioning JSON, CSV and reports prepare configuration for an engineer. Saved projects, virtual tests and published simulations do not configure physical equipment or establish physical commissioning acceptance.