Manual Task Execution

To launch a task, navigate to the TASKS tab. If you have not yet created a new project, you will be working within the demo-village-project by default. This project includes several example tasks that cover a range of system functionalities.

Demo Tasks

The following tasks are included. Note that some require additional hardware beyond the Raspberry Pi itself.

No external hardware required:

  • A minimal task that performs no actions — useful as a blank template.

  • A task that plays sounds when the subject enters a defined area.

  • A task that presents stimuli on a touchscreen and waits for the subject’s response.

Requires Bpod with 3 behavioral ports:

  • A task that delivers water whenever the subject pokes any of the three ports.

  • A task that plays a sound on a center poke and requires the subject to choose one of the two side ports.

  • A task that presents a visual stimulus when the subject is detected within a circular region around a defined point.

Requires Arduino:

  • A task that switches an LED on and off using Arduino.

Training protocol (training_protocol.py):

  • The training protocol contains the following logic: TODO

Launching a Task

Select a task from the list. You can optionally select a subject from the subject selector before running it:

  • No subject selected (None): The task runs with the default settings defined in training_protocol.py. No data or video is saved.

  • Subject selected: The task runs with the settings currently assigned to that subject. Once the session ends, the subject’s settings are updated automatically by training_protocol.py.

Press RUN TASK to start.

Testing the Training Protocol

Clicking TEST THE TRAINING PROTOCOL simulates what would happen if training_protocol.py were applied to a specific subject right now, based on their existing session history. The computed settings for the next session are displayed on screen.

This is intended as a validation tool: if you have modified training_protocol.py, you can use this to verify that the updated code runs correctly against real subject data without making any actual changes. The settings displayed are never saved or applied.