These photocell sockets are standard on a lot of LED fixtures here in the US. Virtually every LED I've seen with a twist-lock PC uses a 7-pin PC socket, aside from early models. Most of the cities and towns in Rhode Island that have gone LED are using wireless nodes made by Cimcon. Pretty neat controls. A little overkill for street lighting IMO and they have lots of glitches and bugs but otherwise a cool futuristic set-up. One of the major reasons the cities and towns here like the nodes so much is because they measure power consumption, so the utility company can bill the city/town for actual usage rather than than estimate. The vast majority of street lights in the US are connected directly to the secondary distribution lines and are not metered, so the utility company charges a fixed rate per street light which incorporates maintenance, estimated energy usage, and of course a generous profit margin, as the utility company essentially operates a monopoly when it comes to public street lighting.

Up until recently in RI, if towns wanted to install their own lighting they could only do so using their own infrastructure. The city/town would have to install their own poles, wiring, and service cabinet (for the meter). A huge expense compared to just connecting lights to the pole transformers. Recently, RI state legislature was passed to allow cities to purchase ownership of the existing utility-owned lights, contingent upon the city/town swapping the lights out for LEDs. This would save the cities and towns money. Not even counting the fact that the LEDs would be using less energy (and will supposedly require less maintenance), the charges from the utility company have been reduced to just estimated power consumption (so about 50% of what the charge per light would normally be). Rather than pay a maintenance charge every month, the city now just pays their own contractor to go around and fix outages. And with the nodes, rather than estimated usage, the cities can be charged actual usage, so they can get the best price possible. That's especially helpful when dimming the lights, as different lights can be dimmed at different times, all while keeping everyone honest about power consumption.

I like the old system best...simple and easy to maintain. We seem as culture to love making things complicated in the name of progress. I like the KISS principle: Keep it Simple Stupid.

I agree. Nothing easier than a simple photocell and an incandescent lamp. Only two components to fail. Even with MV it's simple, as the ballasts rarely fail (unless capacitors are involved). With HPS, you add an ignitor but GE fixtures have come with plug-in ignitors since 1985 and Cooper and AEL both offer the option. With LEDs (or any electronic ballast) there are so many parts to fail inside, none of which are serviceable. One of those little links of the chain breaks and the whole thing is junk. And the nodes, while providing lots of useful information and features, are overkill for most applications.

I agree. Nothing easier than a simple photocell and an incandescent lamp. Only two components to fail. Even with MV it's simple, as the ballasts rarely fail (unless capacitors are involved). With HPS, you add an ignitor but GE fixtures have come with plug-in ignitors since 1985 and Cooper and AEL both offer the option. With LEDs (or any electronic ballast) there are so many parts to fail inside, none of which are serviceable. One of those little links of the chain breaks and the whole thing is junk. And the nodes, while providing lots of useful information and features, are overkill for most applications.

This is one of the biggest reasons why I don't like LED. I like LED if I can use it in an E26 socket...I like my luminaires serviceable.

I've seen a few of the streetlights in downtown Seattle with failed node photocontrols so they are on all day and off all night, while the rest of the lights on the street turn on & off together. The old system of a single photocontrol on a relay box to turn on a group of streetlights made more sense to me.

I've seen similar failures with the node-controlled street lights on the freeways here too. Some dayburn 24/7, some are off 24/7, some strobe at night and run steady during the daytime, and others work dusk-to-dawn independently rather than following the programmed schedule the rest of the lights follow (turn on 30 minutes after sunset, turn off 30 minutes before sunrise). On cloudy days the group-controlled lights won't turn on until it's almost black out, as they come on at a set time based on the astronomical clock instead of based on ambient light level. You can spot the ones running independently dusk-to-dawn, as they usually come on before the rest of them.

RIDOT used to have a mix of individual photocells on freeway lights (older installations mainly, from the 50s, 60s, and 70s) and group-controlled lighting installations (1980s to 2016). When they initially began going LED, the LED lights got PCs if it was an old install and the group-switched ones got shorting caps. Early into the conversion, they decided to use the nodes and went back and changed the PCs and shorting caps on the new LED lights (kept the fixtures) and began converting the rest of the state's freeway lights "en masse", beginning in spring 2016 and finishing by fall that same year. All the relay cabinets for the group-switched lights were bypassed, providing 24/7 power to the poles. The older installs with PCs needed no modification since they had 24/7 power already.

However, the older installs have lots of wiring faults, so some of the LED street lights have never been used ever (and perhaps their HPS predecessors were never used either, depending how old the faults are). Each pole has a 10A fuse in the base. When they went LED many of the poles worked for a short period of a week to a couple months and then the poles went out like they'd been for years. My guess is the faults are water-related, as heavy rain seems to most often knock out some of the lights.