Home | Contents | 日本語

CHAPTER 5

FMPS in Numbers

"If every streetlight has cameras, won't the network, the electricity bill and maintenance costs be enormous?" This is the question we are asked most often. The answer lies in the design: FMPS does not stream footage. The figures below are estimates, to be confirmed by measurement in pilot demonstrations.

Assumptions

Power

In electricity terms: 1.5 W × 24 h × 365 days ≈ 13 kWh, or about ¥400 per streetlight per year (at ¥31/kWh). This is 10–20% of the power of a typical residential LED security light (about 10–20 W).

Where streetlights receive power only at night, the daytime must be covered by a battery: 12 h × about 1.5 W ≈ 18 Wh (about 21 Wh for a 14-hour summer day). Allowing for depth of discharge and conversion losses, a capacity of about two smartphone batteries is needed. Because the battery is cycled daily, long-life lithium iron phosphate cells are assumed.

Storage

Two weeks of storage (about 260 GB) requires a 512 GB high-endurance memory card; a 256 GB card holds about 12 days.

Data volume

Even over an inexpensive LTE Cat.1-class link (about 1 Mbps effective), the images for 1 km of tracking can be retrieved in a little over ten minutes. Where speed matters, a somewhat faster link reduces this to a few minutes. No high-bandwidth backbone is needed.

Cost

It is expensive only because it is not mass-produced

Streetlights with built-in cameras already exist as products. At present, however, a dedicated mould costs more than ¥10 million, and spreading that over sales of a few dozen units a year puts the price at around ¥200,000 per unit. That price reflects small-scale production, not the technology.

Mass production changes the picture completely.

This is what we mean when we say FMPS needs no special new technology: combining two widely used products turns a streetlight into a camera streetlight. Inexpensive LED streetlights have already proven their waterproofing and weather resistance outdoors in very large numbers.

Not demanding 100% reliability

The main reason security equipment is expensive is the reliability demanded of it.

• With "safety" as a product, nothing normally happens. Customers pay for nothing happening.
• Yet if the machine does not work at "that one moment" — once in several years, perhaps once in a lifetime — all the trust built up until then is lost.
• A TV or refrigerator used every day shows at once when it breaks. Security equipment cannot show whether it really works until an incident occurs.

But this applies when you try to protect your own home, by yourself, with certainty. As long as protection rests on a single device, that device must keep delivering 100% reliability, and security equipment inevitably becomes very expensive.

The idea behind the e-JIKEI Network, begun in 2002, is different: everyone watches over everyone. The camera in front of a house is there not for that house but for the public space of the community. Because many cameras complement one another, community-wide watching-over is unaffected even if the failure rate of individual cameras exceeds 10%.

FMPS realises this idea at streetlight scale. The fields of view of neighbouring cameras overlap, so that even if one camera stops, the route is linked by the cameras before and after it. Because no single unit has to be 100% reliable, units can be built cheaply from consumer-grade parts, and no emergency call-out is needed for each failure.

Cost breakdown

• Unit: smartphone-grade electronic parts, about US$100 per unit as a guide (excluding housing, battery and installation). For reference, the camera-equipped LED security light "e-JIKEI Light" announced in 2009 had an expected retail price of ¥50,000–70,000, comparable to the cost of installing a security light alone (Chapter 11).
• Installation: uses the existing poles and power supply. Integrating units when streetlights are periodically renewed (for example, LED conversion) keeps extra work to a minimum.
• Maintenance: redundant design removes the need for a call-out for every failure.
• Operation: the lowest tier of IoT contract is enough for communication. Operating the Recorder and audits adds further cost.

← Previous: 4. How It Works | Next: 6. VRAIO: Governing Outputs →

Contents

• Home
• 1. What Is FMPS?
• 2. A Child Does Not Come Home
• 3. Why Now?
• 4. How It Works
• 5. FMPS in Numbers
• 6. VRAIO: Governing Outputs
• 7. If a Government Tried to Misuse It
• 8. Uses and Phased Introduction
• 9. What It Means for Stakeholders
• 10. FAQ
• 11. Research History
• 12. Contact

FMPS + VRAIO: Public-safety infrastructure built on streetlights, and the governance of AI outputs | Yusaku Fujii (Gunma University) and NPO e-JIKEI Network Promotion Institute
This site presents a research-stage concept; all figures are estimates.