I want to ask if there are other countries that uses computer controlled lighting system, like Israel, as it seems that this mothed of lighting system controlling is an israeli invention.
An example of computer based lighting systems in Israel that I know of their existence includes:
1. Streetlighting of Haifa.
2. Floodlighting of Haifa Ports.
3. Streetlighting of Nesher.
4. Roadlighting of Highway 22 from the center of Haifa to the north of the Kerayot towns.
5. Streetlighting of Jerusalem.
6. Indoor lighting of Castra mall (Including the parking lot lighting).
7. Indoor lighting of Pisgat-Hen elderly sheltered housing (Where my grandma lives).
In the case of roadlighting (1, 2, 4 and 5), most of them are based on Menorah ITC-T streetlighting controllers, that are controlled remotely by a cellular communication from a remote control center with a PC that runs Windows XP and Wizcon apps generator, and the operators are gets the warnings about failures, to their smartphones.

I want to ask if there are similar types of controlled streetlighting systems, outside Israel, because its seems to me, that Israel is considered a pioneering country in this type of lighting system controlling mothed, and that in the rest of the world, the traditional and more reliable mechanical or electronic (but not computer based) photocells, timer switchs, and manual switches (And in case of US/Canada, UK, Australia and New Zealand streetlighting, an independent photocell on each lantern) are used.

I guess the Win (or similar) based computers are already used to run the user interface to control nearly everything around the whole globe, just because it is the cheapest way - you may easily afford quite high degree of redundancy, because the computers and the OS are cheap, allow to easily design even quite complex GUI, so it become intuitive for the operators.
It includes even the high level parts of the "autopilots" for modern high speed trains (the safety things are on most simpler system, but the cruise optimization and such non-safety critical systems systems run on the PC)

So I would be very surprised, when any centrally controlled system won't use the PC's to access the controls.
And over the years the control methods evolved into a form of simpler or more complex web server (even the simple 8bit microcontrollers are capable to run a basic server - e.g. with only one status line and one data entry text-box) running on the target device, so then the UI could be easily written for virtually any computer...

I seen such system working indoors - Basically a very stupid deamon that sits in the services and communicates through EIA232.... Does this even require a PC ? Microcontroller not good enough (and a bit more energy efficient) ?

I seen such system working indoors - Basically a very stupid deamon that sits in the services and communicates through EIA232.... Does this even require a PC ? Microcontroller not good enough (and a bit more energy efficient) ?

The PC's are used mainly for the user interface and some really high level automation.
If the microcontrollers would be more efficient? If all the settings would be stable, then maybe. They are in the target equipment anyway.
But when the difference start is, when you have to control large installations (e.g. lighting in a large city district,...) and the conditions are changing (distribution system load, traffic,...). Then the power saving attainable from more precise control offset the running power of the few PC's. Without them you would have to program the lighting controllers with margin, so lights would stay at full power when the traffic already ceased, so the reduced power would be sufficient. This you can never predict so accurately upfront, but the centralized and convenient PC control allow to respond to that changing situation. Moreover the same PC's function as traffic monitors, so operators control not only the lighting, but as well the junction traffic lights and many other aspects, so on the bottom line you may allow way smoother traffic, so less pollutions from the cars...

The system i know is basically "M - photocell reading = light output level", and this is transmitted to dimmable ballasts. Could as well use a simple controller with a potentiometer or DIP switches to set M,

Same for on/off control, except there you need just to set 2 thresholds Ton and Toff (switch off at darker ambient than switch on, so not just comparator - a controller is really required, but not a PC...)

But to set the potentiometer and/or dip switches, you would need the maintenance crew to travel to each of the control points, open the box, readjust, close it and go to another controller. That mean a lot of traveling - so all that traveling would become way more expensive and energy hungry, so such readjustments would have to be limited. As I do not know any SINGLE case, where such system setting would be done first time in the really optimum way, you would need many visits for the readjustments. Then the conditions (traffic pattern over the day,...) change and you have to readjust again. Without the networked control, nobody would want make any adaptations, so he would set the lights just to operate longer (higher light levels for the photocells, shorter times with reduced power), so the system would operate way from optimal.
If you allow the centralized control (and for that you need all the controllers to be networked), you need only few operators to readjust it nearly daily, so the system would run way closer to what is actually needed, without anyone leaving his chair.
And because these operators usually care for all the infrastructure, the user interface of all of the components have to be nice graphic in order to be intuitive.

And for that the PC platform is the cheapest one, as it's cost is due to high manufacturing volume, even when highly complex, usually fraction of any low volume dedicated controller, regardless how simple it could be. And as it have abu8ndance of the computing power, the required custom software does not have to be optimized for hardware requirement, what would require a huge amount of expert work, so it is extremely expensive.

You still dont need anything more than a simple controller, even for that. Stack it with an Ethernet/X10/whatever networking interface and connect to it from the PC to change settings

The system which i frown upon is powering the PC 24/365.25 so that it does the work instead of conrollers, and all the rest of the system being effectively "computer controlled relays"

I mean the part from where the commands are send...
But even for the field: The PC is the cheapest platform of today, even cheaper than the small board based on a small PIC or ATmega on it...

The main reason is, than for the Atmel or similar you would have to develop specialized software and optimize it, so it would suffice with the tiny computing power available there. The biggest problem then become the security, to protect the infrastructure from attacks from the network. The only method usable for these platforms is the use of separate network, what is very expensive, because the small 8-bit machines could not run the required security software and still be robust against DOS-like attacks to allow the use of cheap open network.

But with the PC you get the operating system ready to take care of all these networking tasks for free, so your only task is to develop the specialized controller part. Moreover you have way more convenient development tools for that, pushing the development cost further down. So with that you could use the cheaply available open Internet connection and still keep reasonable security level.

And today's PC's are available with consumption of 10W and below, so the extra electricity consumption does not cost as much as the specially designed platform.


You should not forget, than there is not so high production volume, where to dissolve the higher development cost required for the simpler microcontrollers, so even when these chips cost less than a dollar, the development make them way more expensive than the PC platforms with plenty of ready to use software components and tools.

Each controller would use a local photocell - so 2 controllres on 2 sides of a mountain would switch each at its optimum time

The logic is still simple and can be implemented in a PIC

The security can be implemented as simple password-based security with suspension time after each failed attempt. Really no need to implement anything hard for it, just compare N bits to a stored value and run a timer if the attempt failed - Pretty much any controller can handle this

Each controller would use a local photocell - so 2 controllres on 2 sides of a mountain would switch each at its optimum time

The logic is still simple and can be implemented in a PIC

It can, but it would require way more effort: design the board, design the enclosures, get them manufactured, program and debug all the code, get the required safety certificates (as it is a product introduction onto the market), you end up easily with few "man-weeks" of engineering work. Now one engineering hour cost about 200..300Euro, you end up way above 20kEuro only to get the design done, so not counting the material (as that is nearly nothing).
When using the PC, you have to pat some 100 Euro for it, but your SW development count in hours (web server controlling a relay) and you are done. So unless you want to install 100's of exactly the same controllers, the ready made PC become way cheaper.

The security can be implemented as simple password-based security with suspension time after each failed attempt. Really no need to implement anything hard for it, just compare N bits to a stored value and run a timer if the attempt failed - Pretty much any controller can handle this

Then you effectively loose the connection in few minutes, simple DOS attack would keep the controller all the time in the "wrong password time-out" mode, so you don't get any chance to access it...

Programmable controller with SELV analog input pins (so you wire just an LDR + resistor to it, forming a divider) and output data pins - check

Relay module SELV --> 240v, tested and approved - check

Ethernet to EIA232 telnet server box - check

Modem with Ethernet output - check

Its all on the market, just connect it all up....



The program for this PIC ? Can write it right here - just convert my pseudocode to the assembly of the correct pic and convert the IFs to JMPs. I dont need any web server, just stupid EIA232 with keyboard input

Program loop cycle would be like this :

Initialization
{
    $wrong_Password = 0
    $its_night = 0
    $light_on = 0
}

Main
{

/* network part */
    If $wrong_password == 0
    {
        Read $input from EIA232
        If $input=="character meaning "log in"
        {
            Echo "Password:" >>EIA232
            Read 50 more characters from input, write to $buffer1
            $buffer2 = FLASH_STORED_PASSWORD
            If $buffer1 == $buffer2 (compare byte by byte)
            {
                Echo "Password:" >>EIA232
                Echo "Access approved" >>EIA232
                $logged_in = 1
                While $logged_in
                {
                    Echo "Enter parameter ID character or log out command" >>EIA232
                    Read $input from EIA232
                    $address = $input
                    if $input == "character meaning log out"
                    {
                        Echo "You are logged out" >>EIA232
                        $logged_in = 0
                    }
                    else
                    {
                        Echo "Enter new value" >>EIA232
                        Read $input from EIA232
                        $data = $input
                        Write $data --> $address in flash
                    }
                }
            }
            else
            {
                Echo "Access denied" >>EIA232
                $wrong_password = (some high value for delay)
            }
        }
        else
        {
            Echo "Input L to log in" >>EIA232
        }
    }
    else
    {
        $wrong_password--
    }

    Read $reset_password from ONBOARD_JUMPER_RESET_PASSWORD
    if $rest_password
    {
        Write "default password" --> $FLASH_STORED_PASSWORD
    }

/* light control part */
    Read $photocell_in from photocell
    if $photocell < FLASH_NIGHT_LOWEST_VALUE
    {
        $its_night = 1
    }
    if $photocell > FLASH_DAY_HIGHEST_VALUE
    {
        $its_night = 0
    }
    if $its_night
    {
        if $photocell > FLASH_DAWN_HYSTERESIS_HIGH
        {
            $light_on = 0
        }
        if $photocell < FLASH_DAWN_HYSTERESIS_LOW
        {
            $light_on = 1
        }
    }
    else
    {
        if $photocell > FLASH_DUSK_HYSTERESIS_HIGH
        {
            $light_on = 0
        }
        if $photocell < FLASH_DUSK_HYSTERESIS_LOW
        {
            $light_on = 1
        }
    }
    Echo $light_on --> RELAY_OUT
}