AI in the grow

[Bandit420]

Thoughts and musings about using AI in the grow room.
If any of us do this I assume like myself, you'll likely go by way of AC Infinity until a competitor jumps in or somebody is willing to spend close to $10K on a commercial set up.

Curiosity and skepticism are at the forefront here for me.
When I started growing indoors everything was analog. Then came C.A.P. fuzzy logic, Titan, Autopilot, and a variety of sensors and triggers to make it work.
Next came basic digital and UIS, and now AI.
Did well when everything was analog. Cant really say anything has made weed better or easier to grow but it has added efficiency and effectiveness of devices plugged into controllers as well as more metrics of information to process for a sometimes more dialed in grow than without it.
It's nuts to think about it but in 20 years we've gone from using analog timers and grow room calculators based off fans' CFM to AI.

Controllers are only as good as the devices that are plugged into them. So it will be very interesting to me if AI can make my devices better than they are or is this just techy useless nonsense designed to get money out of my wallet because it's the new-new?

To start, I'll be using AI to control temps mostly but I wanna see how it learns my room. Am I missing something it will not sorta thing.

I've opted for the lower end controller without CO2.
I already have two CO2 monitors. I really don't see a difference between 450ppm and 1200ppm when it's available in my garden so I'm at the point where as long as I see 400ppm or higher on CO2, that's fine for me and my plants.
Maybe I'll get into VPD control but really heat management comes first over every other metric.

I'll also be using it to monitor media moisture using two probes. One for indica and one for sativa/polyhybrids because when growing two or three varieties together, indicas often take up around half the water of a sativa in the same amount of time in identical conditions so sativas need more water more often. I wanna see how AI "learns" that.

Other bells and whistles I'll add on are wireless dongles, 6 in total, to get rid of some of the UIS wires in my room, and lighting daisy chain dongles to control 2 lights from a single control port.

Maybe I add more later. I could possibly see a use for the peristaltic pump dosing my water tank with nutes or pH, IDK yet.

There are some RJ11 interface cables to control lights with the ACI UIS controllers. You'll have to check your light and brand for compatibility but it looks like most Spider, Mars, Viper and other LED brands coming out of Shenzen, China are compatible.
And as always if you get into this...never pay retail. 420thingsgrow is a coupon code worth 10% at ACI and maybe even a better sale coming for Prime Day in June/July on Amazon

Hydroponics & Growers - CONTROLLERS - AI Controllers and Sensors - AC Infinity

acinfinity.com

 

[OLDCRO]

I already purchased the AI controller going to use it next grow. I am going to let it take over the complete cycle i will hand water of course. Going to let it control fans, light,exhaust fan and if moisture probe is available in Canada by then will try one of those. Using ecowitt probe right now

 

[OLDCRO]

The only problem right now with the moisture probes for the AI controller they have only one plug in for hook up so dont purchase two no place to hook up. Ecowitt with the hub you can hook up 8 moisture probes. Maybe acinfinity will come up with a way to run more than one moisture probes

 

[Tremojem]

I will be using AI in my 2025 Winter grow, as I am growing outside in the Summer until I can upgrade my Lung Room, to actually manage a Summer grow, due to the heat.

I will use all AI, and hopefully be using the wireless methodology. But, admittedly, I am confused as to how ACI is marketing the wireless methodology. $39.00 for each device? $320 to use 8 devices? I don't actually know for certain as ACI is very good at confusing the issue. How many transmitter/receivers are you getting for $39? And, do you need a 1 to 1 for each device? From what I gleaned, you do. So, that is costly if true. It does appear to lead the consumer to believe that you get a 2 pack for $39. So $160 isn't so bad, but...did I really just say that $160 wasn't a bad price for 2 sets of "dongles"???

Regardless, I am excited to use this technology. It's not that I believe in the marketing, or any implied claims, that it will improve my grows. It's convenient. It's another set of eyes. It never sleeps.

But, it won't guarantee results. I'd be a fool to think differently. If set up correctly, it will make my life just a bit less complicated or stressful. I am only speculating so my apologies if I misspoke about an emerging technology that I have only just learned about.

Ultimately, I planned on a second tent, so adding the AI+ to my 69 Pro+, was just smart...kinda AI(ish) in a way.

But, it will be interesting, at the very least. Good to see you started this thread. I hope anyone using the new AI+, and wireless methodology or any combination of either two, join in the conversation and contribute their experiences.

 

[OLDCRO]

Here in Canada you can pre order wireless hookups you get two sets for 55.00 dollars . If in US im sure there cheaper. So for eight wireless hookups in Canada would be a little over 200.00

 

[Tremojem]

Here is an idea. Maybe use the splitter hub. Currently, I use six ports.

Two Circulating Fans
One Exhaust Fan
One Supply Fan
One Humidifier
One LED Light

In this scenario I have no advantage to using the splitter hub but other scenarios may have a real advantage here. Just a suggestion.

 

[Tremojem]

Here in Canada you can pre order wireless hookups you get two sets for 55.00 dollars . If in US im sure there cheaper. So for eight wireless hookups in Canada would be a little over 200.00

Okay, so you made the same assumption as I did, and you get two transmitter/receivers. Cool beans, but still an overcharge for what you're actually getting.

 

[OLDCRO]

Okay, so you made the same assumption as I did, and you get two transmitter/receivers. Cool beans, but still an overcharge for what you're actually getting.

Same tech as wireless mouse and keyboard

 

[Tremojem]

Just finished the pre order for the transmitter/receivers. Already have the AI+ Controller.

 

[Bandit420]

I may have been confused on the sensors. I thought two sensors could be plugged in so instead of CO2 or whatever + moisture I'd want two moisture probes.

My timing is always off so bad with ACI though...Bought the 69 controller and like a week later the 69 Pro plus came out. Bought it, then a few months later the AI came out so I'm always playing catch up.

Yes the wireless to my understanding is two transmitters for $39 USD so I assume that means two devices and if splitter dongles can be used, maybe more.

To control lights needs a whole different type of splitter which they finally came out with so now I won't need 1 port per light and can run two lights off one port.

 

[OLDCRO]

I may have been confused on the sensors. I thought two sensors could be plugged in so instead of CO2 or whatever + moisture I'd want two moisture probes.

My timing is always off so bad with ACI though...Bought the 69 controller and like a week later the 69 Pro plus came out. Bought it, then a few months later the AI came out so I'm always playing catch up.

Yes the wireless to my understanding is two transmitters for $39 USD so I assume that means two devices and if splitter dongles can be used, maybe more.

To control lights needs a whole different type of splitter which they finally came out with so now I won't need 1 port per light and can run two lights off one port.

There is two slots one is for your humidity and temp senser leaving one free. They come out with the moisture senser and the CO2 senser but only one spare slot im thinking they are going to come out with some dongle with multiple hook ups.

 

[Tremojem]

Same here, with regard to always catching up.

I haven't shared this story because I didn't want to sound like an ass, but you all be the judge. At least, I don't remember sharing it lol.

I was in the market for a new LED Light. Money was no object. I already had a cheap Chinese light, and wanted my new light to be on par with the rest of my grow equipment, which is ALL ACI!!! Yup, all of my shit is ACI!

I was doing my research and narrowed it down to three epic choices. During this time and even right before I pulled the trigger, I had three conversations with ACI, about other ACI products. During all three of those conversations I explicitly stated "I am looking for a new LED Light, to replace my current one, and wish that ACI had one" and would then chuckle with them and espoused that I would purchase one immediately if they offered one.

NOT ONE of those CSRs said a word about the new ACI Light that was only a couple weeks away from launch. Additionally, there was nothing on their website citing a pre-order, which is THEIR THING!!!.

I purchased my Photontek 600 W Pro. A week later ACI launched their new line of LED Lights, without a pre-order or any pre announcement.

I was floored.

 

[OLDCRO]

Same here, with regard to always catching up.

I haven't shared this story because I didn't want to sound like an ass, but you all be the judge. At least, I don't remember sharing it lol.

I was in the market for a new LED Light. Money was no object. I already had a cheap Chinese light, and wanted my new light to be on par with the rest of my grow equipment, which is ALL ACI!!! Yup, all of my shit is ACI!

I was doing my research and narrowed it down to three epic choices. During this time and even right before I pulled the trigger, I had three conversations with ACI, about other ACI products. During all three of those conversations I explicitly stated "I am looking for a new LED Light, to replace my current one, and wish that ACI had one" and would then chuckle with them and espoused that I would purchase one immediately if they offered one.

NOT ONE of those CSRs said a word about the new ACI Light that was only a couple weeks away from launch. Additionally, there was nothing on their website citing a pre-order, which is THEIR THING!!!.

I purchased my Photontek 600 W Pro. A week later ACI launched their new line of LED Lights, without a pre-order or any pre announcement.

I was floored.

I also have a Photonteck 600 watt Pro its been running at least three years non stop with no problems great light.

 

[moe.red]

I’ve been struggling a bit with sensor placement. I think that can have a huge impact on how ai learns the room. Aci cables are too short for me out of the box.

 

[Rootsruler]

I've been curious about the AI controller as well. I'll be watching to see what y'all's results will be.

 

[Zen_seeker]

Same here, with regard to always catching up.

I haven't shared this story because I didn't want to sound like an ass, but you all be the judge. At least, I don't remember sharing it lol.

I was in the market for a new LED Light. Money was no object. I already had a cheap Chinese light, and wanted my new light to be on par with the rest of my grow equipment, which is ALL ACI!!! Yup, all of my shit is ACI!

I was doing my research and narrowed it down to three epic choices. During this time and even right before I pulled the trigger, I had three conversations with ACI, about other ACI products. During all three of those conversations I explicitly stated "I am looking for a new LED Light, to replace my current one, and wish that ACI had one" and would then chuckle with them and espoused that I would purchase one immediately if they offered one.

NOT ONE of those CSRs said a word about the new ACI Light that was only a couple weeks away from launch. Additionally, there was nothing on their website citing a pre-order, which is THEIR THING!!!.

I purchased my Photontek 600 W Pro. A week later ACI launched their new line of LED Lights, without a pre-order or any pre announcement.

I was floored.

Similar situation but I didn’t talk to any Photontek reps. Just a lot of searching and store clerk conversations.

I ended up getting the SQ300W Pro on a Black Friday deal. Piece of crap. The switch only does 0%-25-50-75-100%. You need to buy a $460 controller if you want to have full intensity control. In side by side grows my old quantum boards do better and they are not the best quality.

 

[Zen_seeker]

I’ve been struggling a bit with sensor placement. I think that can have a huge impact on how ai learns the room. Aci cables are too short for me out of the box.

For me it’s always fan cords bing about 12” short.

 

[Rootsruler]

After thinking about this for a bit, I'm curious. What will the AI bot learn that you aren't doing already or have automated?

I thought about the processes that I go through and then I put myself in the bots place. What is it that the AI bot could learn and do better that I'm doing now? What and how can the AI bot improve my automated functions?

When I was running CO² in sealed rooms I opted for the C.A.P. Fuzzy logic controller thinking that the fuzzy logic would somehow improve my CO² application and, ultimately, my dry yields. Honestly, I didn't see a difference, even over multiple runs over the older, static monitor the C.A.P. monitor replaced. The static monitor was programmed to open the valve when the room dropped below a setpoint and close once it reached another. So what did the fuzzy logic improve? The only thing I could think of is that the levels wouldn't be quite as spikey but I'm not sure that really matters in the end.

Taking a wait and see attitude to see what this controller can actually improve. Maybe I'm not accounting for other processes that I'm forgetting about but I'm still not sure how the controller can markedly improve what I'm already doing.

I'm using a Controller 67. I really need to upgrade so I really want to see if this is the answer or do I just go with the Pro+WIFI?

 

[Zen_seeker]

Still on the 67 myself. Wait and see…

 

[Bandit420]

After thinking about this for a bit, I'm curious. What will the AI bot learn that you aren't doing already or have automated?

I thought about the processes that I go through and then I put myself in the bots place. What is it that the AI bot could learn and do better that I'm doing now? What and how can the AI bot improve my automated functions?

When I was running CO² in sealed rooms I opted for the C.A.P. Fuzzy logic controller thinking that the fuzzy logic would somehow improve my CO² application and, ultimately, my dry yields. Honestly, I didn't see a difference, even over multiple runs over the older, static monitor the C.A.P. monitor replaced. The static monitor was programmed to open the valve when the room dropped below a setpoint and close once it reached another. So what did the fuzzy logic improve? The only thing I could think of is that the levels wouldn't be quite as spikey but I'm not sure that really matters in the end.

Taking a wait and see attitude to see what this controller can actually improve. Maybe I'm not accounting for other processes that I'm forgetting about but I'm still not sure how the controller can markedly improve what I'm already doing.

I'm using a Controller 67. I really need to upgrade so I really want to see if this is the answer or do I just go with the Pro+WIFI?

Still on the 67 myself. Wait and see…

I use both a 67 and 69pro+.
The only thing you're really missing between the two are more ports at 4 with the 67 vs 8 on the 69pro and yeah that wifi feature for remote control away from the grow.
If you're using 4 or fewer devices and no need to adjust your grow while away then the 67 is fine. You can even expand the ports on the 67 with splitters so then you're only missing out on that remote access.

The AI IMO will only be useful if you install it for everything. I don't believe the baby steps I'm taking into it take advantage at all but maybe later I do.
The advantage of AI will come in making minor adjustments to light intensity in comparison to temp/rh/vpd.
So like say at around hour 8 of lights on your temp has climbed to the point where your plants can only take in so much light before it starts to beat down a plant in combination with excess heat causing stress. If that controller knows to turn that light down and dim it to reduce garden stress and cost of operation, that's pretty cool to me.
But yeah at the end of the day we're really talking minor adjustments that can be beat with a little math and timers if it were analog. You know you're at maybe 75-85% efficiency doing that then you add AI and you could be getting close if not at 100% efficiency.

 

[Tremojem]

Similar situation but I didn’t talk to any Photontek reps. Just a lot of searching and store clerk conversations.

I ended up getting the SQ300W Pro on a Black Friday deal. Piece of crap. The switch only does 0%-25-50-75-100%. You need to buy a $460 controller if you want to have full intensity control. In side by side grows my old quantum boards do better and they are not the best quality.

You can control the Photontek with the ACI Controller

 

[OLDCRO]

I look at it as something new to experiment with if it gives me higher efficiency fine if not fine. Cost wise its not gonna break the bank !!!

 

[Tremojem]

After thinking about this for a bit, I'm curious. What will the AI bot learn that you aren't doing already or have automated?

I thought about the processes that I go through and then I put myself in the bots place. What is it that the AI bot could learn and do better that I'm doing now? What and how can the AI bot improve my automated functions?

When I was running CO² in sealed rooms I opted for the C.A.P. Fuzzy logic controller thinking that the fuzzy logic would somehow improve my CO² application and, ultimately, my dry yields. Honestly, I didn't see a difference, even over multiple runs over the older, static monitor the C.A.P. monitor replaced. The static monitor was programmed to open the valve when the room dropped below a setpoint and close once it reached another. So what did the fuzzy logic improve? The only thing I could think of is that the levels wouldn't be quite as spikey but I'm not sure that really matters in the end.

Taking a wait and see attitude to see what this controller can actually improve. Maybe I'm not accounting for other processes that I'm forgetting about but I'm still not sure how the controller can markedly improve what I'm already doing.

I'm using a Controller 67. I really need to upgrade so I really want to see if this is the answer or do I just go with the Pro+WIFI?

Yup. It was an easy upgrade for me. As I mentioned earlier, I am looking to add a second tent. So, I needed another controller. They offered me a deal on top of a YT discount. So, I would be foolish to not upgrade.

 

[Observer]

It'll be the future with state of the art agriculture.

Ai + Robotics

As will a lot of things....

Imagine a united planet, making software and hardware do all the work for us, while allowing more time to continue to Research, Produce, Innovate and evolve.

Exponential knowledge & Technological growth.

All of these systems are collecting data and being refined.

Compute Power is exploding, and that enables the system to feed its self and grow faster.

Massive Data banks are being developed.

 

[OLDCRO]

Similar situation but I didn’t talk to any Photontek reps. Just a lot of searching and store clerk conversations.

I ended up getting the SQ300W Pro on a Black Friday deal. Piece of crap. The switch only does 0%-25-50-75-100%. You need to buy a $460 controller if you want to have full intensity control. In side by side grows my old quantum boards do better and they are not the best quality.

You can run your light from acinfinity controller you need

AC Infinity UIS Lighting Adapter Type-A . I run a 2000 sqare with this full control​

 

[moe.red]

Ai should be better at maintaining setpoints.

Up to the grower to give it the right environment variables based on how the plants look.

Until there is a camera option and a ton more processing power the value is still based on the grower.

If there were a camera option where daily photos are compared and settings changed based on what is seen that could be groundbreaking. A really advanced ai could see how it’s setpoints make more or less green over time and maximize it for example.

If it could also be used to detect nutrient imbalances that would be neat. That’s pretty tough tho. So much conflicting info on the web it would be using as its data source.

Noticing you did a trim or training and back down the lights and fans to slowly ramp them up would be a nice app.

I’m just thinking that the potential is there but the current iteration is not super helpful to an above average grower. Maybe I’ll change my mind when the hydro probes are released.

 

[Zen_seeker]

I use both a 67 and 69pro+.
The only thing you're really missing between the two are more ports at 4 with the 67 vs 8 on the 69pro and yeah that wifi feature for remote control away from the grow.
If you're using 4 or fewer devices and no need to adjust your grow while away then the 67 is fine. You can even expand the ports on the 67 with splitters so then you're only missing out on that remote access.

The Bluetooth app lets me set things and works from anywhere in the house. WiFi isn’t needed right now. Would have been a game changer when I was working.

The AI IMO will only be useful if you install it for everything. I don't believe the baby steps I'm taking into it take advantage at all but maybe later I do.
The advantage of AI will come in making minor adjustments to light intensity in comparison to temp/rh/vpd.
So like say at around hour 8 of lights on your temp has climbed to the point where your plants can only take in so much light before it starts to beat down a plant in combination with excess heat causing stress. If that controller knows to turn that light down and dim it to reduce garden stress and cost of operation, that's pretty cool to me.
But yeah at the end of the day we're really talking minor adjustments that can be beat with a little math and timers if it were analog. You know you're at maybe 75-85% efficiency doing that then you add AI and you could be getting close if not at 100% efficiency.

Environment is my biggest battle. Most I can usually do is beat it back to a stalemate. Anything lower than 50% is a gift.

Oldcrow has me looking up the UIS again.

 

[Zen_seeker]

You can control the Photontek with the ACI Controller

I was told it wouldn’t work with mine when I got the light. And they weren’t listed as compatible at the time. But Oldcrow had me go take another look just now.

 

[Zen_seeker]

You can run your light from acinfinity controller you need

AC Infinity UIS Lighting Adapter Type-A . I run a 2000 sqare with this full control​

When I got the light I wasn’t happy with the staged dimmer. When I asked about the RJ11 connection I was told it wasn’t supported.

However after seeing your post I’ve taken a second look. It now has Photontek listed but with a caveat.

• PHOTONTEK - X, XT, SQ Series that are compatible with their Dimmer to RJ12 Adapter (Limited Dimming Range)

So I’ll visit my remaining somewhat local store and see what they say. My concern is that I’ll still have to deal with 25% increments. Might still be worth it to get remote control around the house.

 

[Tremojem]

I was told it wouldn’t work with mine when I got the light. And they weren’t listed as compatible at the time. But Oldcrow had me go take another look just now.

It was the first thing I did, when I bought the lamp. I actually ordered the cable from Photontek, in my original order. It was one of the smaller reasons I bought it for...completely compatible with my 69 Pro +. I also purchased an adapter from ACI.