Frequently Asked Questions
- General
- Aren't there other systems that do what IntelliBlinds™ does?
- If IntelliBlinds™ is so advantageous, why hasn't something similar been developed before?
- Has IntelliBlinds™ been independently tested? If not, what confidence is there in the projected energy savings?
- Your savings projections assume that IntelliBlinds™ is used on windows that are over-shaded most of the time. But what about windows that aren't over-shaded, or aren't shaded at all?
- Aren't the periodic shading adjustments associated with Dynamic Daylight Control (DDC) distracting to the building occupants?
- How does IntelliBlinds™ compare with other energy saving devices?
- What is the projected operating life of IntelliBlinds™?
- Does IntelliBlinds™ or its underlying technology have other applications?
- Technology and Design
- How hard would it be to design around the IntelliBlinds™ patents?
- Why doesn't IntelliBlinds™ include wireless networking capability?
- Why doesn't IntelliBlinds™ include wireless remote control?
- IntelliLux™DayLight Sensor (DLS)
- The IntelliLux™DLS senses near-IR wavelengths as a proxy for the daylight level, but don't all lamps produce at least some IR?
- What happens if there is incandescent lighting in the room?
- Aren't near-IR wavelengths affected differently than visible wavelengths by passage through the atmosphere and reflection from room surfaces?
- How does the output of the IntelliLux™ DLS relate to photometric quantities like illuminance and luminance?
- If the output of the IntelliLux™ DLS isn't correlated with established photometric quantities, what does the DDC function of IntelliBlinds™ actually regulate?
- What is the response time of IntelliBlinds™ in making DDC adjustments? How does this affect its ability to block glare?
- Why doesn't the IntelliLux™ DLS need to be calibrated like conventional WorkPlane Illuminance (WPI) sensors?
- How will IntelliLux DLS work with next-generation LED-based lighting systems?
- IntelliLux™Artificial Light Sensor (ALS)
- The IntelliLux™ ALS senses artificial light by detecting the flicker produced by AC or pulsed lamp current. Are there any lighting systems that don't produce enough flicker to be detected by the IntelliLux™ ALS?
- Can the signal from the IntelliLux™ ALS be used for more than just sensing the on/off state of the lighting system
- Market Potential
- Your market projections refer to "demand-limited" sales. What does "demand-limited" mean, and how do demand-limited sales relate to potential actual sales?
- Have you done any test marketing? If not, what confidence is there in your sales projections?
- Why are you targeting just the non-residential daylighting market?
- Your market research emphasizes retrofit applications in existing buildings. What about new construction?
- If the potential market is as lucrative as you say it is, why aren't similar products already available?
- If the potential market is as lucrative as you say it is, won't competitors jump in to dilute your projected sales?
- Your sales curves suggest that it will take almost 15 years to achieve peak annual sales. Why? And isn't that too long to fully exploit the market gap?
- Commercialization
- How do you plan to get IntelliBlind to the market?
- If IntelliBlinds™ is so great, why aren't you manufacturing it yourself?
- What kind of company would be best positioned to commercialize IntelliBlinds™?
- What needs to be done to put IntelliBlinds™ into high-volume production?
- Wouldn't IntelliBlinds™ need a specialized distribution network to achieve anywhere near the sales you project?
- Are there opportunities for bundling IntelliBlinds™ with related products?
- How would IntelliBlind be maintained and serviced?
General
Aren't there other systems that do what IntelliBlinds™ does?
Yes, there are other systems that perform Dynamic Daylight Control (DDC) for non-residential daylighting applications, such as those described in The Conventional State of the Art—but at no less than three times the cost of the IntelliBlinds™ Model D. And all other DDC systems require integration with remote sensors, control computers, and lighting systems, creating major issues for distribution and installation. So, while other DDC systems are available, the IntelliBlinds™ Model D is the only DDC system that's inexpensive and simple enough for large-scale commercial deployment.
If IntelliBlinds™ is so advantageous, why hasn't something similar been developed before?
We believe there are four reasons we were the first to be able to develop a DDC product with the game-changing advantages offered by the IntelliBlinds™ Model D:
- We thoroughly researched the market before embarking on IntelliBlinds™ development. Surprisingly (to us, at least), this doesn't seem to be typical of technology-intensive new-product development. See Market Research for more information.
- Unlike many would-be developers of DDC products, we don't have a vested interest in an existing product line or technology, which allowed us to take a clean-sheet-of-paper approach to IntelliBlinds™ R&D.
- We have technical expertise that is relatively unique in the automated shading and daylighting industries. Our technical strengths are in control algorithms, optoelectronics, microcontrollers, mechatronics, systems engineering, and user-system interface design—all of which were important in developing IntelliBlinds™.
- And, finally, IntelliBlinds™ wasn't developed overnight; it took 10+ years of sustained R&D. Other would-be developers of similar products may not have had the resources or patience for such a long-term effort.
Has IntelliBlinds™ been independently tested? If not, what confidence is there in the projected energy savings?
No, IntelliBlinds™ has not yet been independently tested. But now that the technology and design have been refined to our satisfaction and the associated intellectual property fully protected, we intend to subject IntelliBlinds™ to a real-world test program administered by a Government or University Lab. See Research and Demonstration Program for more information on what we have in mind.
However, there is relatively little doubt that the IntelliBlinds™ Model D will deliver the savings we project (see Energy Savings and Payback Periods). Our projections are corroborated by published data from independent sources, including university laboratory testing of an experimental system that is functionally equivalent to IntelliBlinds™.
Further, the savings we project are the median savings across the entire U.S. market. In certain areas (such as those with high electricity costs and high average insolation) and in certain buildings (such as those with high installed lighting power density), the savings will be significantly greater than our projections. Early adopters could gain an extra measure of confidence in the savings provided by IntelliBlinds™ by limiting initial use to just such high-payoff installations.
Your savings projections assume that IntelliBlinds™ is used on windows that are over-shaded most of the time. But what about windows that aren't over-shaded, or aren't shaded at all?
Extensive research over the past three decades has shown that most windows that receive enough daylight to make daylight harvesting worthwhile are chronically over-shaded. See The Need for Dynamic Daylight Control for a discussion of that research.
But it's true that some non-residential windows aren't over-shaded; we estimate the fraction of such windows as 25%. So our market projections for IntelliBlinds™ exclude the 25% of the potentially applicable floor area to account for windows that aren't chronically over-shaded. This is based on the assumption that, in a given building, IntelliBlinds™ wouldn't be used on the facade with the lowest average base-case shading (typically the north-facing facade for buildings in the U.S.).
Further, our savings projections are based on a distribution of base-case shading settings that takes into account the fact that some of the remaining 75% of windows are also relatively un-shaded.
And, finally, to further reduce the uncertainty in the expected savings in any given project, an assessment of the base-case shading settings could easily be performed as part of the pre-retrofit building survey. This could be done either through photography of the building's facades, or as part of the initial walk-through. IntelliBlinds™ could then be omitted on any facade that doesn't exhibit significant over-shading.
Aren't the periodic shading adjustments associated with Dynamic Daylight Control (DDC) distracting to the building occupants?
In a word, no. Our DDC algorithm minimizes the frequency of shading adjustments, and testing shows that people quickly become acclimated to them. Further, the subjective appeal of increased daylight—coupled with the ability to adjust the daylight level as desired—fosters occupant acceptance during the brief acclimatization period.
How does IntelliBlinds™ compare with other energy saving devices?
The metric typically used to evaluate energy saving devices is the payback period. The IntelliBlinds™ Model D has a projected typical payback period of just 2–4 years, which is competitive with energy saving technologies that have already achieved high market penetration.
And because the payback period of conventional DDC technology— at least 8 to 10 years—is far too long to achieve substantial market penetration, IntelliBlinds™ won't face any direct competition in the marketplace.
What is the projected service life of IntelliBlinds™?
We are still in the process of optimizing the the IntelliBlinds™ Model D design and support plan to maximize return-on-investment to the end-user. Currently, we are targeting a Maintenance-Free Operating Period (MFOP) equal to the median payback period (about three years), with a service life of three MFOPs.
Does IntelliBlinds™ or its underlying technologies have other applications?
Yes. Most of the IntelliBlinds™ technologies and design are also highly advantageous for the residential automated shading market, in which convenience—rather than energy savings—is the primary product benefit. We have prototyped and tested residential versions of IntelliBlinds™, and have thoroughly researched the residential market (which is potentially far larger than the non-residential energy savings market). However, the residential market is harder to penetrate, requiring a much larger investment in manufacturing and marketing. Accordingly, we don't plan to aggressively pursue the residential market until after IntelliBlinds™ gains a foothold in the non-residential market.
Technology and Design
How hard would it be to design around the IntelliBlinds™ patents?
The IntelliBlinds™ patent protection (in force and pending) is quite broad, so it would be difficult to develop a non-infringing device that exploits the same innovations.
Of course, it may be possible to develop different innovations that would a enable a non-infringing DDC device as simple and inexpensive as the IntelliBlinds™ Model D. But from our own experience, we know that developing such a device—if feasible at all—would take many years.
Why doesn't IntelliBlinds™ include wireless networking capability?
The current Evo-3 prototype does, in fact, include wireless networking capability in order to facilitate testing, demonstrations, and daylighting research. And wireless networking would likely be included in a residential version of IntelliBlinds™. However, the current baseline production design does not include wireless networking capbility, for at least three reasons:
- Achieving a short payback period is crucial for success in the non-residential energy savings market. While wireless networking technology seems inexpensive, the incremental cost of adding it would increase the average payback period by a full year.
- Networking capability is useless without other systems and software to take advantage of it. While integrated building control systems are becoming more popular, they are still relatively rare, so additional costs (beyond the incremental costs of adding the capability to IntelliBlinds™ itself) would be incurred in taking advantage of networking. These additional costs would increase the payback period by another several years.
- Finally, and most importantly, the IntelliBlinds™ Model D omits networking capability because—unlike all other DDC systems based on conventional technology—it simply doesn't need it. Other DDC systems require interfaces to other devices, but IntelliBlinds™ performs DDC autonomously, eliminating the need for wired or wireless interfaces.
Of course, wireless networking capability could easily be retained when the current Evo-3 prototype is productionized, if so desired.
Why doesn't IntelliBlinds™ include wireless remote control?
Wireless remote control capability would certainly be included in a residential version of IntelliBlinds™. The baseline production design currently omits wireless (IR or RF) remote control for two reasons:
- Achieving a short payback period is crucial for success in the non-residential energy savings market. While wireless remote control technology seems inexpensive, the incremental cost of adding it would increase the average payback period by several months (for IR-based remote control) or nearly a year (for RF-based remote control).
- Our testing and market research show that, in non-residential buildings, the perceived value of remote control is substantially lower for automated shading systems capable of DDC operation, because DDC eliminates much of the need for deliberate shading adjustments. And this is true to an even greater degree for IntelliBlinds™, which incorporates IntelliState™ automation to automatically anticipate occupants' shading preferences when DDC isn't needed.
Of course, wireless remote control capability could easily be added if it becomes advantageous in the future.
IntelliLux™ DayLight Sensor (DLS)
The IntelliLux™ DLS senses near-IR wavelengths as a proxy for the daylight level, but don't all lamps also produce at least some IR?
Yes, virtually all lamps produce detectable amounts of IR. In fact, incandescent lamps radiate most of their power at IR wavelengths, which is why they are so inefficient. But the near-IR produced by high-efficiency lamps—such as fluorescent and LED types—is negligible compared to the near-IR from the daylight in a typical daylit space.
What happens if there is incandescent lighting in the room?
There shouldn't be any incandescent lamps in buildings in which energy efficiency is important. But if there does happen to be incandescent lighting in a room in which IntelliBlinds™ is installed, IntelliBlinds™ will interpret the incandescent light as daylight. In practice, however, this is not a problem, for two reasons:
- Virtually all of the incandescent lighting still in use in non-residential buildings is task lighting, not area lighting. If any illumination from incandescent task lighting reaches the IntelliLux™ DLS, it will generally be insufficient to cause a significant bias in the DLS output.
- If enough incandescent light does reach the IntelliLux™ DLS to affect its output, the result will be a slight reduction in the actual daylight admitted by IntelliBlinds™. This can be compensated for by twisting the IntelliTwist™ Wand to regain the desired daylight level. So, unless the incandescent lighting is frequently switched on and off, there will be no significant impact on the average daylight level (and, hence, the average energy savings).
Aren't near-IR wavelengths affected differently than visible wavelengths by passage through the atmosphere and reflection from various surfaces?
Yes. For example, a clear blue sky is dark at near-IR wavelengths, but may appear quite bright to the eye. And the reflectivity of certain surfaces (e.g. healthy vegetation) can differ substantially between near-IR and visible wavelengths.
However, those differences actually favor the use of near-IR as a basis for closed-loop daylight control. Specifically, the ratio of the power at near-IR wavelengths to the power at visible wavelengths is much greater for sunlight than for diffuse daylight. So the IntelliLux™ DLS is much more sensitive to sunlight than are conventional illumination sensors.
And that's a major advantage, because sunlight—whether directly from the solar disc or reflected from nearby glass facades—is the major cause of glare in daylit spaces.
How does the output of the IntelliLux™ DLS relate to photometric quantities like illuminance and luminance?
Under a fixed set of sky, ground, and window-shading conditions in a given installation, the output of the IntelliLux™ DLS could be modeled as a function of the daylight components of the window luminance, the wall luminances, the horizontal illuminance, and the vertical illuminance on a plane parallel to the window surface—all measured at/from the center of the daylit zone. But this is a complex function that varies in an even more complex way with changes in weather, season, position of the sun, and window-shading setting.
However, the relationship between accepted photometric quantities and the subjectively perceived daylight level (and particularly the perception of glare) is poorly understood, so there would be little benefit to developing an accurate model of the IntelliLux™ DLS output as a function of such quantities. Much more important is the fact that the IntelliLux™ DLS output is reasonably well-correlated with the subjectively perceived daylight level, and (especially) that it is highly sensitive to perceived glare.
If the output of the IntelliLux™ DLS isn't correlated with established photometric quantities, what does the DDC function of IntelliBlinds™ actually regulate?
The purpose of DDC is to maximize the average level of useful daylight, thereby maximizing the savings from daylight harvesting. But it can't do that if it alienates the building occupants. And the two surest ways for a DDC system to alienate building occupants are to (1) allow glare within the daylit space, and (2) make excessively frequent (and thereby annoying) shading adjustments. On the other hand, office occupants can comfortably tolerate a relatively wide range of daylight brightness.
So tightly regulating the daylight according to some established photometric quantity isn't just unnecessary, it's also undesirable because it requires frequent shading adjustments. Instead, an ideal DDC system would just keep the shading as open as possible without risk of excessive brightness or glare. And thanks to the IntelliLux™ DLS, that's exactly what IntelliBlinds™ does.
What is the response time of IntelliBlinds™ in making DDC adjustments? How does this affect its ability to block glare?
Response time is a consideration for DDC systems because it represents the potential duration of transient high-glare conditions (e.g. when the solar disc emerges from behind a nearby building, causing a sudden increase in the illuminance on the window).
Indeed, response time is a serious issue for DDC systems using ElectroChromic (EC) windows, which can take minutes or longer to achieve a significant change in visible transmittance. Some sort of predictive control system is probably necessary to achieve acceptably short glare durations with an EC-based DDC system.
On the other hand, motorized blinds typically have a limit-to-limit slat-tilt cycle time of less than 10 seconds, resulting in a typical DDC response time of the order of 1 second. This is short enough to be a non-issue except in rare circumstances.
And it's even less of an issue for IntelliBlinds™, because its IntelliLux™ DLS sensor can detect incipient glare even at the edges of the daylit area. That means that IntelliBlinds™ typically starts closing its host blind well before glare reaches the occupants, reducing the glare duration to near zero.
Why doesn't the IntelliLux™ DLS need to be calibrated like conventional WorkPlane Illuminance (WPI) sensors?
When performing DDC, the IntelliBlinds™ Model D is intended to loosely maintain whatever perceived daylight level has been set by the user—not to maintain some specified absolute photometric (e.g. luminance or illuminance) quantity. Therefore, absolute accuracy isn't important; rather, what matters is that the output of the IntelliLux™ DLS is reasonably well-correlated with the perceived daylight level and highly sensitive to glare. And our testing shows that is indeed the case.
How will the IntelliLux™ DLS work with next-generation LED-based lighting systems?
As long as a lighting system doesn't radiate appreciable power at IR wavelengths—which is true for all high-efficiency lighting systems, including those based on LEDs—the IntelliLux™ DLS will work with it.
IntelliLux™Artificial Light Sensor (ALS)
The IntelliLux™ ALS senses artificial light by detecting the flicker produced by AC or pulsed lamp current. Are there any lighting systems that don't produce enough flicker to be detected by the IntelliLux™ ALS?
In the U.S., virtually all area and task lighting systems produce flickering illumination that is readily detectable by the IntelliLux™ ALS.
However, as is the case with indandescent bulbs (such as those used in flashlights), it is possible to drive an LED with a pure DC waveform, so LED-based lighting systems producing flicker-free illumination are feasible. But generation of a pure DC waveform from the AC mains voltage is cost-prohibitive, and control of the drive current (which is necessary to prevent burning out the LED) is done most efficiently using a switching regulator. Further, efficient dimming also requires a pulsed waveform. So, because high-frequency flicker is essentially invisible to the human eye, and because smoothing the pulsed drive current to pure DC before applying it to the LED adds substantial cost, few (if any) LED-based area lighting systems can be expected produce completely flicker-free illumination. And the IntelliLux™ ALS is sensitive enough to detect even the slightest flicker in ambient illumination.
Can the signal from the IntelliLux™ ALS be used for more than just sensing the on/off state of the lighting system?
Potentially, yes. The current Evo-3 prototype uses the IntelliLux™ ALS output to track the dimming level of the lighting system for research purposes. However, that requires a lighting-system-specific calibration, because the absolute amplitude of the illumination flicker doesn't always vary monotonically with the brightness of the lamps. This and other uses for the signal are currently the subject of in-process patent applications.
Market Potential
What is the practical significance of your market projection?
Our Market Projection is for demand-limited sales. Demand-limited sales are the sales that would be achieved if enough manufacturing, advertising, and distribution capacity were in place to reach every buyer willing to purchase the product. If the demand-limited sales projections assume a retail price high enough to yield an industry-standard profit per unit, and if the product has IP protection—both of which apply to IntelliBlinds™—then the projections provide a realistic measure of the commercial value of the IP.
Further, because investments in manufacturing and marketing such a product will yield relatively high returns, the probability of actually achieving the demand-limited sales is fairly high—provided that the commercialization is appropriately resourced and executed.Have you done any test marketing? If not, what confidence is there in your sales projections?
Test marketing is generally regarded as the best way to gauge a new product's sales potential. However, we have not yet done any test marketing, for two reasons. First, only recently has the market evolved to make commercialization of a product like the IntelliBlinds™ Model D a highly lucrative, low-risk proposition. And second, we have been focusing on refining the underlying technology, with several patent applications currently in process.
However, pending test marketing, our analytical sales projections (see Market Potential) are actually quite conservative, for at least three reasons.
- The projections are based on conservative estimates of IntelliBlinds™ cost and energy savings (and, hence, payback period), and assume no increase in electricity prices or non-residential square footage going forward.
- The projections are limited to retrofits only, and exclude new construction and major renovations.
- The projections assume zero growth in the aggregate non-residential floor area.
Further, because demand for products in the non-residential energy conservation market is strongly driven by a single tangible benefit, demand is relatively easy to predict analytically: the ultimate market penetration of an energy saving product for non-residential buildings is highly correlated with the inverse of its payback period. For a product that provides a projected average payback period shorter than a few years, ultimate market penetration is invariably high—assuming that it doesn't annoy the building occupants and is easy to distribute. And that's certainly the case with the IntelliBlinds™ Model D.
Why are you targeting just the non-residential daylighting market?
The non-residential daylighting market is much easier to penetrate than the residential automated shading market, with far lower costs of advertising and distribution.
On the other hand, the residential automated shading market is substantially larger than the non-residential daylighting market, and the residential version of IntelliBlinds™ (without DDC capability, but with wireless remote control and networking capability) shares more than 80% of the design and components of the non-residential DDC-capable version. For this reason, while our commercialization plan initially focuses just on the non-residential market, we will certainly target the residential market as non-residential sales and production capacity grow.
Your market research emphasizes retrofit applications in existing buildings. What about new construction?
Our market projections exclude sales for new construction and major renovations purely out of conservatism.
Like most other energy conserving products, the IntelliBlinds™ Model D costs even less in new construction and major renovations than in retrofits. In fact, over the product's life-cycle, new construction and major renovations will account for far more sales than retrofits.
However, the retrofit market is much more lucrative in the near-term—but typically much harder to penetrate—than the new construction market. Therefore, ability of IntelliBlinds™ to penetrate the retrofit market is commercially very significant and well worth emphasizing.
If the potential market is as lucrative as you say it is, why aren't similar products already available?
Other products that perform the same function as the IntelliBlinds™ Model D—i.e. dynamic daylight control for daylighting applications—are, indeed, commercially available. But because of high cost (and, hence, excessive payback period), their market penetration is very low (and effectively zero in retrofit applications).
The need for a cost-effective dynamic daylight control system has long been recognized, but such a product just isn't feasible with conventional technology.
If the potential market is as lucrative as you say it is, won't competitors jump in to dilute your projected sales?
They will certainly attempt to do so. But because it took many years and multiple patent-protected innovations for us to achieve the cost-effectiveness necessary to penetrate the market, it's unlikely that competitors will achieve substantial success in the near-term. And if competitors do eventually succeed in designing-around the patents, IntelliBlinds™ will have already gained a dominant market share.
Commercialization
How do you plan to get IntelliBlinds™ to the market?
We're currently evaluating several commercialization approaches, but our current baseline plan is to license the IP to another entity.
If IntelliBlinds™ is so great, why aren't you manufacturing it yourself?
That is certainly an option that we are exploring fully. Due to the simplicity of its design and the fact that its sales projections are based on ample mark-ups, IntelliBlinds™ could be contract-manufactured in low volume and still sold at a profit, which could then fund higher-volume production.
What kind of company would be best positioned to commercialize IntelliBlinds™?
Ideally, the IntelliBlinds™ Model D would be commercialized by a company that:
- has an established presence in the non-residential energy efficiency market, and
- has the ability and willingness to commit the resources required for commercialization, and optionally
- is involved in manufacture or distribution of complementary products.
Examples of such companies include Energy Services Companies (ESCOs) and manufacturers or distributors of daylight-harvesting lighting controls and/or occupancy sensing light switches.
However, due to the relative ease with which IntelliBlinds™ can be commercialized, companies meeting even just one of the aforementioned criteria would enjoy a high probability of market success. Manufacturers of shading products or residential appliances, for example, might also be well-positioned to commercialize the IntelliBlinds™ Model D.
What needs to be done to put IntelliBlinds™ into high-volume production?
Relatively little. The existing third-generation prototype design is nearly production-ready, although another design iteration would be advantageous for very high-volume production. See Development Status.
Wouldn't IntelliBlinds™ need a specialized distribution network to achieve anywhere near the sales you project?
No; the IntelliBlinds™ Model D is easy to distribute and fully compatible with the existing distribution network for energy efficient lighting controls.
Are there opportunities for bundling IntelliBlinds™ with related products?
Because the IntelliBlinds™ Model D is compatible with virtually any miniblind and any daylight-harvesting lighting control, it requires no bundling and can be distributed independently. For example, it could be retrofitted to buildings which already have miniblinds and daylight-harvesting lighting controls, providing an extremely short payback period.
However, because few non-residential buildings are already equipped with daylight-harvesting lighting controls, the more likely scenario is that IntelliBlinds™ will be bundled with at least a lighting control (and probably other energy conserving lighting products) as part of a comprehensive lighting efficiency retrofit program. In addition, it may be advantageous to include miniblinds in the bundle for the relatively few buildings not already equipped with them.
In such a bundling scenario, the individual products could simply be co-distributed without any design changes—or the designs could be optimized for greater synergy. For example, minor changes to miniblind design and firmware for both IntelliBlinds™ and the lighting control could significantly improve daylighting performance. We are in the process of filing patent applications in this area; contact us for more information.
How would IntelliBlinds™ be maintained and serviced?
We have not yet finalized a recommended support plan. However, support would be facilitated by the following factors:
- While we have not yet completed a detailed analysis of Mean-Time-Between Failures (MTBF) for IntelliBlinds™, we expect an MTBF no lower than that of HVAC convectors and lighting system components.
- Due to the high expected MTBF and small size, buying and keeping enough spare IntelliBlinds™ units on hand to support even a large building would add little to the building-averaged payback period and would require relatively little space.
- The IntelliBlinds™ firmware includes on-board diagnostics functions, and installation/removal of IntelliBlinds™ requires no specialized expertise. This, coupled with the high expected MTBF, means that faulty units could be identified and replaced with on-hand spares through the same routine inspections that building engineering staff use to identify and replace faulty lamps.