Light Is the Hardest Constraint in the House

Of the prohibited categories of work on Shabbat, kindling is the one explicitly named in the text, and it is also the one modern life has embedded into every surface of a building. A switch on the wall, a motion sensor in the hallway, a refrigerator door, a car dome light, a phone screen, and a status LED on a coffee maker are all the same problem wearing different clothes.

Two points shape every engineering decision that follows. First, extinguishing is prohibited as well as kindling, so a product cannot solve the problem by turning something off. Second, the prohibition is not limited to incandescent filaments. The classic analysis treats a glowing filament as literal burning, but mainstream halachic authorities also prohibit completing a circuit to light an LED, on separate grounds. An engineer who assumes solid-state lighting sidesteps the issue has misread the problem.

The market is concrete: hundreds of thousands of Orthodox households concentrated around Brooklyn, Lakewood, Monsey, the Five Towns, Baltimore, Chicago, and Los Angeles, plus the institutions serving them. They buy lighting, appliances, and building controls, and they filter every purchase through this constraint.

Motion Sensors: The Problem Everyone Runs Into First

Occupancy sensing is now standard in exterior security lighting, closets, garages, hallways, and increasingly in commercial energy codes. It is also the most common Shabbat failure in a modern home, because walking down your own hallway turns on a light. Even where authorities differ on culpability, the practical result is that people avoid the space. Four design responses, in increasing order of sophistication.

  • A hard disable. A physical switch or a mode setting that removes power to the sensor and either leaves the load off or forces it continuously on. Crude, cheap, and by far the most reliable. It must be settable in advance and must not require any interaction during Shabbat.
  • Scheduled bypass. The fixture holds a location-based candle-lighting and havdalah schedule and automatically enters a fixed state for the duration, the capability discussed in Shabbat timers and controllers, moved inside the fixture instead of a plug-in module.
  • Continuous operation with scheduled changes. The light stays on at a set level and changes only at preprogrammed times. Energy code officials accept this less readily, which is worth checking during specification.
  • Indirect-action designs. Architectures where the user's presence does not cause the change but at most removes an inhibition from a process running on its own clock. This is the grama approach, legitimate engineering but also where marketing claims most often outrun rabbinic acceptance.

A caution on that last category. Products marketed as halachically permissible switches have generated serious public disputes, with some certifying bodies endorsing an approach and others rejecting it. Do not build a business on a mechanism you have not run past a recognized authority in writing.

The Mechanical Shabbat Lamp Still Wins

The most widely accepted solution in this whole area is not electronic at all. A Shabbat lamp is a fixture with a rotating opaque shade or a sliding shutter. The bulb stays powered continuously; the user rotates the housing to block the light when the room should be dark, and rotates it back to release it. No circuit is switched, no filament is lit or extinguished, and nothing is measured or controlled.

It is worth studying as a design lesson. Instead of adding intelligence to satisfy a constraint, it removes the electrical event entirely and solves the user's actual need, which is control of illumination in a room, not control of a circuit. The engineering that remains is unglamorous and real: thermal management for a bulb enclosed by a shade, detent durability over thousands of rotations, light leakage at the shutter edge, UL listing for a fixture that intentionally traps heat, and a mechanism a child cannot break.

The principle transfers. A mechanical iris over a display, a sliding cover over an indicator cluster, or a blocked light pipe accomplishes what no firmware mode can, and never needs an opinion about whether a microcontroller state change counts.

Rethinking Every Indicator on the Product

This is where most product teams get caught, because the lighting question is not only about lamps. Any product sold into this market has to account for every element that changes visibly in response to a user action: a power LED that lights when a lid closes, a charge indicator, a backlit button, a display that counts down, a screen that wakes on touch.

Three options that work, in order of preference. Make the indicator state-independent, holding a fixed state through the whole period regardless of what the user does, the standard approach in certified Sabbath mode appliances where displays freeze and tones are suppressed. Move the feedback to a non-electrical channel: a mechanical flag, a detent the user can feel, or a printed position marker, since good indicator design often finds the LED was never the best answer. Or physically obscure it with a shutter or an opaque cap over the light pipe.

What does not work is dimming the LED to near zero or slowing a PWM duty cycle. The circuit is still being modulated by the user's action, and the issue is the action, not the brightness.

Where US Codes Collide With the Design

American designers face a constraint their counterparts elsewhere may not. Egress and emergency lighting is governed by NFPA 101 and UL 924 and must operate automatically on loss of normal power. Commercial energy codes in many states mandate automatic shutoff and occupancy sensing. None of these can be defeated for a Shabbat mode, and a product that offers to disable them will fail inspection and expose the building owner.

The workable approach is scope separation: keep life-safety circuits fully automatic and untouched, and apply Shabbat behavior only to general and task lighting on separate circuits. Say this explicitly in the installation instructions, because the electrician is often not the person who understands the requirement. Institutional projects are where this gets complicated, and the sequencing questions are covered in synagogue automation.

Certification Is the Product Requirement

In this market a technical solution without a hechsher is not a solution. Buyers look for an endorsement from a recognized body, most often Star-K or the Zomet Institute. Engage them early: their review shapes architecture rather than approving a finished design, and a rejected mechanism late in development is a full restart. The process and timelines are described in halachic product certification.

Plan the whole system rather than the fixture. Lighting rarely lives alone in a modern home, and the interaction with thermostats, door sensors, and voice assistants is where the edge cases hide, as laid out in kosher smart home automation.

Build Lighting the Market Will Actually Accept

Projects House develops lighting and control products for observant users: sensor bypass architecture, scheduled operation with accurate location-based times, mechanical alternatives where electronics cannot help, indicator strategy across the whole product, and a certification path agreed with the authority before the design freezes. Send your product concept through our contact form.