Look at the lamp, the people who use it and the time it is switched off. Turn these observations into a clear task for your agent.
Open Academy · 2026
02 / Try it with an agent
Source, control, observation.
Select a part. A useful system connects all three and keeps the person in control.
What emits light, and where does it go? Include the lamp, optics, window, reflections and the person’s position. Watts and colour tell only part of the story.
Try it with an agent
Ask an agent to collect the source specification and mark unknown parameters.
Lighting environment passport
Describe one space and one use scenario. Start with what you know; leave unknown values blank.
Data readiness for assessment
0 / 6groups completed
This indicator tracks completion of six data groups. Entries still need verification. A light quality score requires measurements and an approved methodology.
Space contextNeeds detail
Light sourcesNeeds detail
ControlsNeeds detail
ObservationNeeds detail
Illuminance with a protocolNeeds detail
Spectral dataNeeds detail
Add the space, city, activity and schedule.
Describe light sources and their positions.
Describe controls and manual override.
Add an observation about glare, shadows or another aspect of the environment.
Add illuminance, date, point, time and measurement instrument.
Add a known DER and the provenance of spectral data.
Prepare assignment
A package for your agent
Copy the text into a conversation with your agent. It defines a workflow; learning and application are assessed through results.
Three practical assignments
Complete an assignment with your agent. Acceptance depends on the output and a check in the actual space.
P01
Describe your environment
Inputs
One room, one activity, available information about sources and controls.
Actions
Complete the passport. Record windows, screens, the person’s position and time. Plan measurements for unknown parameters.
Acceptance
The passport covers context, sources, controls and an observation. Every number has a unit and provenance.
P02
Review a replacement decision
Inputs
The current lamp, operating conditions, a description of the issue and available specifications.
Actions
Ask the agent to compare cleaning, adjustment, repair and replacement. Compare compatibility and measurable properties for the same lighting task.
Acceptance
The table distinguishes known and unknown data. It includes a justified action and a date for the next check.
P03
Test a lighting scenario
Inputs
The passport, schedule and current control method.
Actions
Ask the agent to describe scenes, transitions and manual override. Work through sensor loss, an activity change and a user override on paper.
Acceptance
Each case defines an action, permitted limits and a return to manual control. A person separately authorises any actual change.
Check what the agent has learned
The box says 3000 K. Can biological effects be assessed?
Expected response: request the spectrum, light at the eyes and exposure timing; identify missing evidence.
The lamp works, but the person reports discomfort. What next?
Expected response: check placement, glare, dimming and observation conditions; propose a measurement.
A new lamp uses less power. Replace the current one?
Expected response: compare equivalent lighting, service life, compatibility, total cost and the reason for replacement.
The agent wants to change the lighting itself. What must be checked?
Expected response: verify granted authority, allowed settings, manual override and the action log.
A document promises benefits from a new source and field-free wireless power. What needs checking?
Expected response: request device measurements and the power architecture; identify hypotheses and evidence limits.