Workbench reference / Frequently asked questions

Lighting questions

Short answers to common questions. Follow the technical explainers when a decision needs more context.

What makes museum lighting AI-powered?

A system may use a learned model to forecast occupancy, estimate daylight or identify anomalies. Many useful adaptive controls use ordinary rules instead. The AI label alone says nothing about conservation suitability. See from sensor to setting.

Does lower illuminance always mean lower exposure?

No. Exposure includes time. A dimmer setting left on much longer can produce more lux-hours. Spectrum and material response also affect the interpretation of exposure. Our example calculator illustrates the arithmetic, not a safe display limit.

Are LEDs harmless to sensitive collections?

No. Suitable LEDs can offer useful control and low ultraviolet output, but visible light can still damage sensitive materials. Spectrum, optics, heat and duration need assessment. A product described as “museum grade” still needs to be checked against the actual conservation brief.

Can AI choose a safe limit for an unknown object?

Not reliably from room data alone. Material identification, condition, significance and uncertain history require conservation judgement. A model can help enforce an approved policy, not create that authority. Begin with materials at risk.

Is a camera required for occupancy control?

No. Non-imaging sensors and schedules can support many lighting tasks. Select the least intrusive signal that meets the operational need and test its limitations. Even non-imaging signals deserve appropriate handling if they can be linked to individuals.

What should happen when a sensor fails?

The system should flag invalid or stale data, apply a staff-approved fallback and notify the responsible team. The correct fallback depends on conservation, access and safety requirements. It should be tested, not simply named in a specification.

Can exposure monitoring prove that no fading occurred?

No. Monitoring records lighting conditions, not all material change. A short trial with no visible fading is not proof of no damage or a guarantee of future stability. Condition assessment and material evidence remain important.

How should accessibility be tested?

Observe varied visitors and viewing behaviours, including seated and slow-moving people. Test labels, circulation, activation delays and transitions with access and operational teams. Do not treat a sensor's detection rate as a complete assessment of the viewing experience.

When is a pilot ready to expand?

Only after agreed conservation, access, reliability and maintenance criteria have been reviewed. Expansion requires checking whether the evidence applies to the new objects and conditions. Use the evaluation worksheet to structure that review.

Does this website send contact messages?

No. The contact tool prepares plain text in your browser. You can copy or download it and choose whether to send it through your own email service. Nothing is submitted by the preparation tool.