Home The System The Atom The Economy The Opposition The Cost

How to eliminate 95 percent of crime

Before proceeding further, there are issues to be addressed.

The first is an individual's right to privacy. Legally and constitutionally, the system can only be used to investigate reported crime. The private detective down the street cannot legally access the system. There is no more risk to having one's privacy invaded than exists now. If someone wants to invade your privacy now, there is nothing stopping them from doing it.

The second issue is protection from one's own government.

The system must be controlled at the county level, so that the county board can vote ordinances regarding its use and to prevent its misuse, or even vote the system out of existence if need be, should the government start fining people $50 for driving 31 in a 30 mph speed limit zone, or attempts in some other way to use the system as a club with which to unfairly extract money from it's citizens or rob them of their constitutional rights.

Although the main power for the system will come from sixteen Federally controlled 1000 megawatt nuclear reactors specifically built for this purpose, all the rest of the equipment will be owned by the individual States, so as to leave no doubt as to who has the ultimate jurisdiction and control over the system. We don't need an electronic implementation of McCarthyism, and we don't need the people who gave us Waco and Ruby Ridge having control of the system. Access, perhaps, but not control.

Although it will take a coordinated effort by the States at the Federal level to get the system installed and operational, the Constitution clearly defines the role of the State and Federal governments, and leaves no doubt that the ultimate jurisdiction and control over such a system falls to the individual States

Now that these details have been dealt with, we can proceed with the description and details of the system.

The basic concept is to use the immense and inexpensive digital storage capacity of the new laser holographic memory and biochemical memory technologies to store ultraviolet, standard light, and low light level camera output, storing two month's worth of output from each camera at any given time, from a network of microminiature video cameras that covers every square inch of the entire United States, except of course the inside of people's houses. Individual businesses will decide if they want to totally or partially incorporate the system inside their establishments.

This concept is no larger and no more complicated than would be a project to duplicate the existing U.S. telephone network, or the existing cable network, or the existing power line network. It is therefore not a difficult project, since three nearly identical projects have already been undertaken, and successfully completed.

If handled correctly, economically it will take less of a percentage of the gross national product than taken by the Apollo program to land man on the moon.

But it is an immense project. And, additionally, sixteen 1000 megawatt nuclear reactors must be built to provide electrical power for the system. A separate page for this issue, accessible below, will correct misconceptions about nuclear power, and give far more details on this aspect of the project.

If the government were to contract this project out to the private sector, the cost would be so prohibitively expensive that you would be considered a fool for even suggesting it.

However, this project can be economically feasible, if labor from the prison system is utilized to do almost all of the manufacturing (fiber optic cables, cameras, digital multiplexers, computer boards and chassis, interface stations, laser holographic memory, etc...), and (supervised) lesser offenders from the community corrections system do most of the installation, in order to keep costs down.

This is poetic justice, I might add...

Done this way, the costs are entirely acceptable, and are realistically estimated at most to be 160 billion dollars. This is 80 billion for fifteen 1000mW light water nuclear reactors and one 1000mW combination light water / liquid metal fast breeder reactor, and 80 billion for the camera network. Actually, these are bloated estimates which allow for some degree of pork barrel economics.

It will be shown later that the system will save the taxpayers 160 billion dollars per year.
While it might cost a public utility $10 billion to build one 1000mW light water reactor, this cost is highly inflated by legal battles involving opposition by ignorant local landowners and oil and coal industry competition.

The federal government will not face the high costs of organized labor unions, legal battles (since all 16 reactors will be built on land the Federal government has already gone through the legal process of approving and setting aside for the permanent long term storage of nuclear waste), and tax and licensing and consulting fees. In addition, prisoners will do most of the concrete work and most other non-critical construction, and DOD reservists with applicable technical experience will do most of the rest of the work. These reservists include current employees from the likes of Bechtel Corp. (reactor design engineers) and Dresser-Rand Industries (turbine design and manufacture), etc...

Large corporations are more than willing to donate used manufacturing equipment.

Intel corporation still has the equipment used to manufacture the 80386 microprocessor. This equipment has been made obsolete by the advent of the i486 and Intel Pentium technologies, and will be melted down and made into toasters soon. Intel will not sell this equipment, because to sell it might allow a potential competitor to begin operations.

However, this equipment, obsolete as it might be for the purpose of manufacturing top of the line, state of the art microprocessor chips, is far more than adequate for the manufacture of charge-coupled semiconductor devices, which are the heart of the detector component in a video camera.

For the immense public relations value alone, Intel would be more than happy to donate or borrow this equipment, and perhaps even lease some of their manufacturing facilities for free or at cost. There are a large number of other corporations similarly motivated to participate, who have the equipment and facilities to apply to the task of the manufacturing.

Click here for a description of the system's operation, or keep reading for the technical, detail, and construction descriptions, which are followed by the operational details.

The hideously immense (yet inexpensive) storage capacity of laser holographic memory will allow 2 month's worth of surveillance, from 3.2 billion cameras, in standard NTSC video (digitized) format, to be stored initially, with later expansions to 6 months, 1 year, and 2 year's worth, and later expansions in the number of cameras.

There have been working prototypes of laser holographic memory since 1989. The device uses a laser to write microminiature holograms into a crystal or array of crystals.

(The system would work equally well with 20 inch diameter laser disks (20 inch diameter CD-ROMs), but it would cost 50 million dollars a year to resurface / recycle the laser disks. In this case, the systems only saves the taxpayers $159,950,000,000 instead of $160 billion...)

Storage capacity: 1000 2 hour long movies, stored in digital laserdisk quality, in a package the size of a pack of cigarettes. Non-volatile, read-write memory. (Memory is retained when the unit is powered down, like core memory). The access speed is so fast it can directly replace computer RAM.

This technology is the most appalling development yet in the field of digital technology. If every person who has ever lived on the face of the Earth had video surveillance covering them 24 hours a day, every day of their lives, a package the size of the Empire State building could store all of the recorded video, in standard NTSC video format, with room to spare.

It won't hit the open market anytime soon, because that part of the government which is actually capable of thought, and in rare cases, intelligent thought, has seen that the economic damage caused via laser holographic memory's destroying every corporation that manufactures VCR's, audio tapes, video tapes, floppy disks and drives, hard drives, computer RAM, etc... and likely a terrible stock market crash, is far too high a price to pay at the present time.

Over time, it will be phased into the marketplace, and eventually you will be able to buy a 20 terabyte cartridge to replace your computer hard disk and computer RAM for a hundred dollars. But it might be another 10 or even 20 years before you see this happen. The first decent article I read about it was in Byte Magazine in late 1989, October or November if I remember correctly, and there have been many other articles since.

The government is probably using the technology to meet the vast memory requirements of artificial intelligence experiments.

To refresh your memory, standard NTSC video is the video broadcast standard used in the USA for the broadcasting of color television. Every TV set in the United States conforms to this standard. US citizens cannot watch Canadian television, unless they buy a TV set in Canada, because Canada and Great Britain use the PAL broadcast system. Russia uses the SECAM system. These are the three main worldwide broadcast standards. They are not compatible unless you have a special converter.

NTSC uses 30 distinct screens per second, with each screen divided into two parts which are sent separately as 60 frames per second which are interleaved 2 to 1 at the receiver to reproduce the original 30 screens per second. This method reduces monitor flicker and the eyestrain it causes. The resolution of each screen is 525 lines (504 lines are visible, the other 19 carry status information) of scan, with each line having 400 or more pixels (dots of light).

NTSC uses a vertical scan rate of 60 cycles per second and a horizontal frequency of 15.525 thousand cycles per second. These refer to the x/y coordinate of the projection point on the screen (which pixel is being illuminated). Alternatively, the NTSC transmission method can be used as non-standard NTSC to send digital information rather than video, at a rate of 12 million bits per second. Usually, in standard NTSC, the first 19 lines of scan per frame carry other than video information. If you felt that there was a need, you could modify standard NTSC so that 29 screens per second were sent, with digital information replacing the 30th screen, and a standard TV could receive this signal with minor (but noticeable) degradation of picture quality. The anti-crime system uses standard NTSC, because of certain manufacturing cost advantages, but with certain liberties taken and modifications made.

The cameras will be fixed focus color cameras, with the focus set to infinity, as a film camera can have it's focus set to infinity. Everything beyond 10 or 12 feet is in focus this way. These cameras will also be fixed-field, that is, there is no altazimuth mount, no pan or zoom capabilities. The cameras always have the same field of view.

These will be arranged in arrays of three cameras, enclosed in a tunnel enclosure to protect the lenses from the elements. The tunnel enclosure will have a wing attached, so that if the wind speed rises too high, to the point where dust or rain can be blown onto the camera lenses, the wing will rotate the entire tunnel enclosure downward and out of the wind. The cameras inside the mount will then be useless, but since every area will have double coverage from opposing directions, the area of coverage will always be scanned. When the wind dies down, the unit will rotate back into its normal position.

Ultraviolet (preferably) or infrared cameras are used to allow the computer network to automatically track suspect's locations and movements at 10000 times normal speed. It is easier to lose one's shadow than to escape one's own ultraviolet or infrared output... From a known location 2 months in the past, the system can track a suspect to where they disappeared into a house in seconds. The computer then looks for someone exiting the house, and shows the operator the images of all those who left, until the suspect can be visually reacquired.

An ultraviolet (or infrared) camera can be enhanced two ways for use in this system.

Multiple receiver elements, each sensitive to a different range of ultraviolet frequencies, can be installed in the camera, alternating service per screen as the main receiver inside the camera.

30 individual CCD receivers can easily be etched on each charge coupled device chip, and a 30 facetted lens and frequency filters can be used, so that a digital switching controller alternating between each of the 30 CCDs provides 1 complete scan from each device, and so 30 screens per second standard NTSC video output, but with each screen being generated by a different CCD. Each of the 30 screens will be ultraviolet, but in a different frequency range.

This way, if some fool assumes they can disappear into a crowd, or enter a heavily populated building or area, disguise themselves, and leave undetected, they will be in for a rude awakening.

Each person produces a distinct and fairly unique ultraviolet frequency distribution, reliable and unique as a fingerprint. This frequency distribution changes from hour to hour, but not from minute to minute or from second to second.

Although it would be difficult or impossible to track an individual through a dense crowd by using a single ultraviolet frequency scan range, using 30 different ranges makes for a certain target lock no matter how dense the ultraviolet output from a given area. This technique can be reliably used by the computer to track an individual through the most densely populated crowd.

Even with the UV changes that occur over the span of a week, however, this method of identification would still be more than 80 percent reliable. This can be used to search automatically for an individual leaving a crowded area that lacked camera coverage who produces approximately the particular frequency pattern witnessed before the suspect entered the building or area, and drastically reduce the number of false identifications made before the suspect is found again.

The other enhancement is to take advantage of the fact that certain ultraviolet frequencies will travel much further through water than most other UV and visible light frequencies. Whereas the human eye can only discern objects 100 feet away during a hurricane, or 10 feet away during a blizzard, the cameras will be able to see 2 orders of magnitude further in these weather conditions. So, some of the frequency filters used in conjunction with the 30 facetted lens in each UV camera will scan these frequency ranges.

Each camera enclosure will have three cameras. These are a standard NTSC video camera, a low light level sensitivity NTSC camera capable of seeing a candle flame at 1000 feet, and the 30 element ultraviolet camera. Infrared may be used in place of ultraviolet, but ultraviolet is preferable because ultraviolet light travels further than infrared through water.

Only two of these three cameras will produce output at any given time. Each camera enclosure will have only two video outputs. One of these continual outputs is from the ultraviolet camera.

A digital switching circuit will automatically chose between the standard and low light level camera outputs, depending on the output signal of the low light level camera. At night, an occasional standard light camera output screen may be inserted from time to time by this circuit into the stream of low light camera output screens, the switching circuit merely temporarily switching between the two camera's outputs. All three cameras will be in NTSC synchronization, with a single horizontal and vertical sweep generator circuit shared between them, so that this screen interleaving is possible. This also makes the scanning computer's job easier and less complicated, when the system needs to scan through memory to find a given location's screens and switch between screens stored from different camera stations.

For example, at night, the low light camera will be in use, but car headlights may cause the switching circuit to alternate and interleave the screens between the standard camera and the low light camera. At most this would produce 15 screens per second of standard light output, alternating with 15 low light output screens. The light from a gunshot at night would produce, as output, one or two screens of standard light during the stream of mostly low light level screens.

This alternating output is a definite advantage, because the computer system only needs the ultraviolet camera output to track a suspect, but a human operator needs something visually identifiable, and with screens from both light and low light cameras, it is certain that screens suitable for identification purposes will be produced.

The average high school student could program the tracking system. The images are stored exactly as received, and indexed together in the laser holographic memory array. The known 3D coordinate values of the edge of each screen are stored permanently, so the search program can easily step from screen to screen as a suspect leaves the field of view of one camera bank and enters that of the next. Each camera bank's field of view will slightly overlap that of the surrounding banks, and there will be at least double coverage in all areas.

The search program need be no more complicated of that of a laser guided missile.

Basically, the program reads:

There is no optical recognition involved, only simple numerical analysis.

Assuming (very accurately) that the main execution loop takes 350 Intel Pentium machine instructions, each taking an average of 3 clock cycles to complete, and assuming an efficient multitasking far read-ahead large static RAM cache being used to prefetch screen data from the holographic memory array, with the data being delivered from remote storage through a dedicated 100-Base-T or high speed fiber optic network, 158095.2380952 program loops or screen steps can be completed in a single second on an IBM compatible Intel Pentium 166mHz computer.

Since there are 30 screens per second, this equates to the machine being capable of stepping through data while remaining locked onto the center of a specified ultraviolet light source at a rate of 5269.841269841 recorded seconds per second of real time, or approximately 5000 times the speed of real time. It is closer to 5000 due to the computer's interrupt system's processor cycle demands. On a dual Pentium-166 processor machine, this rises to a scan rate of 10000 times the speed of real time. Note that the images will not be displayed on the operator terminal while the computer is stepping through the data.

So, an off-the-shelf IBM compatible dual 166 mHz Pentium processor machine is quite capable of acting as a system access terminal. Motherboard manufacturers would beg to allow their buying power, motherboard circuit board etching negatives and facilities, and night shift assembly lines to be used for the manufacturing in return for the public relations value alone.

If Compaq Computer corporation does this, they can turn around and advertise on television afterward: We are Compaq, and this is what we are doing to save your butts... (although not exactly in those words)... the public relations value could easily double Compaq's sales volume.

There are many methods available to network the cameras, interface terminals, and individual storage subprocessing stations together.

Each will have to be subjected to extensive power consumption, manufacturing and parts replacement costs, reliability, expansion compatibility and potential, and bandwidth capacity testing to determine the best approach.

For example, because of the high cost of copper and the low cost of nuclear power generation, certain remote areas in the countryside or forest, where there are far fewer cameras than in a city, might be better suited to pure fiber network cabling, with the power to the cameras being fed right through the cable itself in the form of light, with a solar cell converter inside each camera bank to convert the light received to usable electricity, and the digitized camera signals flowing in the other direction through the cable at a different light frequency. Newer low power consumption semiconductor technologies make this possible. There is still a large power loss, since this is a very inefficient method of power delivery, but in remote areas with fewer camera units, the low cost of nuclear power allows one to get away with some degree of waste.

In most areas, the power will be delivered in a more standard manner, through copper wires in the form of electricity. Some of the camera units will have battery backup, and some will not.

Here, the main question is which type of video signal output is best. Where in the network should the digitizing take place? There are many different approaches using existing technology. Each camera bank's output might be digital, or it might be analog, with the A/D conversion done elsewhere. Different types of areas would require different approaches to obtain the lowest power consumption and lowest costs,

Almost certainly, in cities and towns, multiplexers will be used to collect and combine the outputs of many individual camera banks, and convert these to one combined output through a beam chopper and out a fiber optic cable that leads to a storage substation, where the holographic memory is located. The exact equipment that will be chosen depends on many factors, which will be determined through further research.

A standard combination of 100 Base-T ethernet and digital fiber uplink output multiplexers would work perfectly, using off the shelf equipment.

Certainly, the only components that needs to be designed are the laser holographic memory storage units and their associated multiprocessor banks, and the minor design details outlined above regarding the ultraviolet or infrared CCDs and lenses. Almost all the rest can be done with off the shelf equipment.

The Federal government has already set the legal precedent that in cases of national security interests, the Federal government may use anyone's patent rights to their heart's content without compensation, or with minor compensation as the government may dictate.

I would think that significantly reducing the possibility of another World Trade Center or Federal Building being blown up would fall into the category of a legitimate national security interest.

However, it is highly unlikely that the government will have to exercise this power, because of manufacturers' willingness to participate without compensation. There are large numbers of these companies and corporations, and many of their plants sit idle during the night shift. Most of these companies and corporations would willingly allow their designs and assembly lines to be used for free or at cost.

These include computer manufacturers, chip and other component manufacturers, steel mills and machine shops, copper mines, copper wire makers, coaxial and fiber cable makers, plastic injection molding facilities, and others, and for the nuclear plants, turbine makers, pipe makers, gauge makers, concrete production facilities, etc...

These companies and corporations will not be blind to what the economic improvements associated with this system can do for them, or to the social and other values and benefits of the system, or to the public relations value.

The system can easily be designed to be fail-safe. Each camera bank, not at first, but through future expansion after the initial start up, will have at least three lines of communication, and three separate power line feeds.

Each camera bank's outputs will be networked to three separate output multiplexers. In this way, three output lines, and three different networks carry three identical copies of the output signals from a given camera bank (although only one copy will be stored, at least at first). If a subnetwork fails, two others will provide backup. If a power line fails, two others will provide backup.

To reduce system start up costs, these backup facilities will be implemented, not immediately, but in later system upgrades.

Also, this setup provides the capability to instantly detect and locate the exact position of any interruption of the system's signals. The speed of light is a constant, and can be used to precisely calculate distance. With three independent signals flowing through each line or cable in the system, each to a different location, if the line or cable is cut, the precise location of the break can be instantly located, and an operator instantly notified and given a live picture of the location, because all areas will have at least double coverage. In cities and towns, there will be quadruple or quintuple coverage.

In future expansions, the extreme cold of space allows the use of supercooled ultraviolet detectors capable of tracking a rat on the street through a Category 5 hurricane. High speed ultraviolet scanning satellites can backup the system in the near impossible case of system failure. The resolution of these satellites does not need as high as a top of the line spy satellite, capable of reading a license plate from orbit, because only the capacity to distinguish an ultraviolet light source and track it are required.

An interesting thought: if the system failed, who would know that it had?

Another: if the system were only 50 percent reliable, would it be an improvement over what we have now?

The laser holographic memory substations would be the only weak points in the system, and would be heavily armored and protected from attack, including subterranean attack. Each would be surrounded by electrostatic detection fields, each would have internal and external ultraviolet proximity sensors, and be completely fenced in, with both solar and diesel generator power backup. It is likely the generator backup systems will be installed as an upgrade, to reduce initial costs.

This covers the preliminary main system technical and design details, and most of the minor details.

The system in operation:

Most people are not stupid enough to commit a crime if there is no hope of getting away with it.

But let's say that such a fool actually existed, and go through a scenario involving a shoplifter at a Wal-Mart.

Wal-Mart begins taking a weekly inventory as soon as the system becomes operational.

During an inventory, an item proves to be missing. An operator accesses the computer / camera / memory system and enters the date of the last inventory, and calls up the video surveillance taken at that time, and uses a mouse and cursor to click upon the correct aisle, and then clicks upon the general location of the item on the shelf, which calls up a more detailed view, and then clicks on the item's exact location, which gives the computer a 3D coordinate to scan. There are cameras at the ends of the aisles and in the ceiling, and so there are 3D x/y/z coordinates.

The computer, at 10000 times the speed of real time, continually scans all ultraviolet activity at the specified 3D coordinate (a hand went to that location on the shelf) for the entire week.

Within seconds, the operator is presented with a sorted list of occurrences.

The computer has linked with the cash register computer log, and checked for all occurrences of a hand going to the shelf without the item being purchased at the cash register within 75 minutes of the hand being detected, and placed these highlighted entries at the top of the sorted list, in a group separate from the other entries in the list. A spectral analysis is done by the computer to check the UV output frequency distribution of the hand to see if the person with that UV fingerprint bought the item at the register. Since a person may have picked up two of the same items at once, pocketed one and bought the other, these majority of entries that indicate a legitimate purchase are not discarded, but are placed at the end of the sorted list.

There are three possibilities with the other entries which do not fall into that sorted category. A shopper may have picked up the item and examined it, and decided against purchasing it, and replaced it on the shelf. The computer will have checked for two detections within 30 seconds, and within 60 seconds of each other, and analyzed via frequency distribution fingerprinting. These entries are placed lower and separate in the highlighted list than the more likely entries.

On a really bad luck day for the operator, out of a list of 2500 entries, and a highlighted list of perhaps 200 entries, 150 of these would likely fall into this category. So, the computer places the remaining most likely 50 highlighted entries at the very top of the highlighted list.

A shopper may have picked up the item, and set it back down elsewhere in the store, or someone stole the item.

It is 99.9 percent certain the one of these first 50 entries in the list is the correct entry. It is 99.95 percent certain one of the first 200 entries in the list is the correct entry, and 100 percent certain that the correct entry is present somewhere in the entire sorted list.

Since the computer knows the precise 3D coordinates of every shelf in the store, the computer can go through each entry in the list, and for each entry, the operator can place a cursor on the suspect, search forward in time, and within five seconds provide the operator a still picture of every occurrence of the suspect in question's hand coming within proximity of a store shelf or fixture. It would take an average of about 1 minute of operator time to step through all these in succession, manually looking for a possible destination of the item.

Another method is to instruct the computer (by clicking a toolbar or pulling down a menu choice as in Microsoft Windows) to lock a subject's ultraviolet and track them through time to where they exited the store or arrived at the cash register. Once these two time indexes are located, the operator can have the computer divide the time, and be shown a picture of the suspect which was taken at a midpoint time index exactly between the time of picking up the item and arriving at the exit or register. Either the suspect still has the item, and this can be seen, or they do not.

If they do, split the time index again forward, and if they do not, split it backward, to arrive at a time index either 1/4 or 3/4 into the total duration of the two original (picking up and exiting) time indices. The suspect is manually examined again to see if they have the item or not, and the time index is split forwards or backwards again depending upon the operator's determination. This process is repeated until the computer determines that 20 seconds or less of unexamined time remain between the last two reference indices, and the operator then instructs the computer as to which of the two 20 second videos he prefers to see, before or after the current time index, depending on whether the suspect still can be seen with the item or not. This entire location process would take about 1 minute, plus the 20 seconds of video. This must be repeated for every entry on the list, but in the worst case scenario, this would take at most 2 hours. There won't be many thefts to investigate, once word gets out regarding the futility of such attempts. This is on par with current average security camera scan times, where store security must watch entire videos.

There is another sophisticated technique that can be used, which involves the fact that an item will always absorb some amount of UV radiation from a customer who picks the item up, and will re-radiates this at the same frequency distribution pattern of the customer, for some time afterwards. The item itself can be tracked using this technique.

Let's say it was a normal day, rather than a bad day, and the original highlighted list only contained 25 entries, and a shoplifter was seen pocketing an item after only 4 other suspects were inspected.

In this case, the Police are notified by the Wal-Mart operator, and the Wal-Mart operator clicks on the image of the shoplifter exiting the store to tell the system which ultraviolet source to lock on to, and uploads the lock and image data to the Police operator's computer terminal.

At 10000 times the speed of real time, the main system tracks the shoplifter out into the parking lot, into their car, and on down the street to the first location they disappear into which does not have any internal surveillance.

Let's say that the suspect entered an apartment building, and went into one of the apartments, the last scene being sent from the apartment hallway cameras. The Police operator calls up the address database and interfaces to it, and assuming the location's data is present in the database, instructs the computer to watch all the known exits from the apartment (doors, windows, even chimneys). If the database does not contain the exit data, it is then entered into the database manually by the Police operator, by using the mouse and toolbar to bring up different views of the location, and marking each exit manually by tracing it's outline by selecting a corner and dragging the mouse to expand an area selection rectangle on the screen until the exit is completely covered, and then releasing the mouse button. This is the same rectangular area selection technique common to Microsoft Windows based paint, graphics, drawing, and CAD/CAM programs. Circular and obelisk area selection patterns are also usually available.

At 10000 times the speed of real time, the computer then watches the exits and keeps scanning the recorded surveillance frames one after another, until an ultraviolet light source is detected. This means that someone walked past one of the possible exits, or someone actually entered or exited. The computer continues to scan forward in time.

At this point, 25 seconds have elapsed for the Police operator since the initial upload from the Wal-Mart operator, if the location was present in the database. If the operator had to manually enter the data, 4 minutes have passed.

5 to 10 seconds later, the operator is presented with a list of every ultraviolet presence at every exit, sorted in their original time index order (first first, last last).

A mouse click on an entry in the list brings up the still video of the exit where the detection took place, at time index plus four seconds from the initial ultraviolet detection (it takes time to open a door).

Since color NTSC video cameras recorded the surveillance, the scanning processor can actually do some fair degree of color recognition. Simple constructs such as the bottom half of the light source's counterpart light or low light level screen read mostly blue, and the top half red, allows the computer to correlate the ultraviolet and standard or low light screens made at the same time index, and be able to detect on a basic, but not truly intelligent level, that the suspect was wearing blue jeans and a red shirt at the time of their disappearance into the apartment.

This capability is only good for sorting the list presented to the operator. If there is an good basic color match, the computer can place these entries at the top of the list of exit activations it presents to the operator.

At any rate, the operator can go through these entries first, and if the suspect is not visually recognized, then the operator goes through the rest of the time sorted index list.

If the suspect is reacquired, the initial process is repeated and repeated, following the suspect until they disappear into an area that has no surveillance, waiting until they come out, and again until the time index of real time is reached, and the Police send someone to the current location to pick the suspect up.

If, after manually checking all these entries, the suspect is not reacquired, there remain three possibilities. Either the suspect is still inside in real time, the suspect left wearing a disguise, or the suspect escaped through a hidden exit that had not been covered originally.

The first and third possibilities are easily enough investigated with a search warrant. If a hidden exit is found, the exit's other end is plotted by searching the location, and that exit is added to the list of known exits for that location in the database, and the entire Police operator process is repeated from the time index of the last disappearance.

If the suspect escaped wearing a disguise, the above-described spectrographic analysis is used to plot the ultraviolet output frequency spectrum of the suspect in the 30 standard ultraviolet frequency ranges.

One second of stored NTSC scan, that is, 30 frames, are analyzed by the computer for light output level in each range. Remember, the ultraviolet camera produces 30 screens per second, but each of the 30 screens covers a different UV frequency range, because of the 30 CCD elements and 30 facetted lens in each UV camera. Each of it's 30 frames per second output is generated by a different CCD element, and covers a different frequency range.

Of course, the longer the suspect stays inside before exiting, the more deviation their UV output will have from that which they radiated at the time they had entered, but if a large number of people exited the location and the span of the time index was, for example, 6 days, the spectrum analysis would still prove a valuable time and resource saving tool for the Police, saving hours or even days of unproductive searching until the correct suspect was reacquired.

In the worst case, there would be a large time frame, and a large number of people who exited the location, a private boxing match for example. In this case, it might take 4 or 5 days using 5 operators to reacquire the suspect, but the suspect would be found again. There is no escape from the system, once it wants to find you. It sees everywhere it needs to see to be able to find you. You would have better luck trying to escape your own eventual death, or trying to escape from gravity. You could disappear into an Indian Reservation, but it isn't likely you would succeed in escaping, unless you have more money than the government to pay the Native Americans there, or actually have more money and never leave. Besides, a year or two after the initial system startup, there would be UV satellite upgrades in orbit, which could not be escaped.

There are also methods to provide for border and shore attempts at escape, but since space is limited here, and this is complicated, these will not be described. Later, these methods will be augmented by on demand satellite surveillance, with the location of every UV source that crosses a border or leaves a shore continually being updated. The satellites need not cover vast areas. They only need to be able to scan at very high speeds, checking a known location, recording UV source movement, checking the next location it needs to check, and so on and so forth, checking and updating the locations of hundreds of thousands of individual UV and infrared light sources per second. This technology already exists, and is in operation now.

The navy has ship and submarine detection and tracking bouys; radar and IR satellites exist to detect aircraft.

The penalties for such attempts at escape, into large private crowds, across borders, or across shorelines, would be somewhat more severe than normal.

There exists the possibility of someone who has legitimate access getting into the system, and programming it so as to give false surveillance. It is expected that this will undoubtedly go on to some degree, but there are limits to how far one can go without considerably endangering themselves. A government with its hand into the lottery where it ought not to be, can probably successfully avoid being caught printing up winning scratch off tickets for their nephews, but if there are any crimes against persons or property, it will be immediately noticed that someone has tampered with the system, when no perpetrator can be found. In this case, help from an outside area can be called in, and anyone who had legitimate access will lose their employment and be prevented from ever occupying that position again. In an area where there is a pattern of such occurrences, other government agencies may randomly add satellite surveillance to the general area, and this will almost certainly catch the perpetrators in the act. Sometimes there will be random satellite scans without any prior notification in randomly chosen areas, and the amount of this will be considerably expanded over time. The penalties for this type of activity would be very severe, perhaps even carrying the death penalty, so that there is further strong deterrent to crimes against persons and property, and the waste of extremely expensive satellite resources.

Considering this, it is unlikely that crimes against person and property will occur on other than an extremely rare basis.

The main system, and most of the details, have been adequately described here.

I could describe some of the truly minor details, such as lens cleaning. Because of the tunnel design of the camera enclosures, it is likely the lenses will only have to be cleaned once every five to ten years. Crews on continual duty can arrange the cleaning schedule so that the complete rotation is completed every five to ten years. The expected labor costs for this activity are far less than 100 million a year, which is nothing compared to the overall economic benefits of the system.

Other minor details need not be discussed further here. This page is under construction, and if any intelligent issues, more specific data, or important omissions are brought to my attention, I will update this page. Please leave intelligent comments through the entry form on the home page.

I have pointed out that there are economic advantages associated with this system. Now that you have more respect for the seemingly impossible, Follow the economics link below to examine these advantages in complete detail. Drug dealers and users, however, may wish to visit the Opposition page first, for some enlightening and likely surprising information.

Home The System The Atom The Economy The Opposition The Cost

Copyright 1997, 1998, Robert J. Nelson.

See also COPYR978.GIF, an integral component of this work, for a full legal copyright notice.

copyright notice

This work is composed of the following copyrighted files: COPYR978.GIF, FIXCRIME.HTM, SYSTEM.HTM, NUCLEAR.HTM, ECONOMIC.HTM, OPPOSE.HTM, and COST.HTM. No other files are covered under this copyright.

You must have the author's express written permission to reproduce this work if either of the following are true:

1) It is part of any package sold for profit, or if you gain monetary profit from the use or publication of any part of this work.

2) It is your intention to republish this work, or if you publish this work, in any medium which contains advertising of any kind.