Ally Monk and Ramen Dutta co-founded Motorleaf, an indoor hydroponic and greenhouse automation startup in 2015, when Ramen made the first working prototype (image of a current product below). CEO Ally shared with us the history of Motorleaf, the role of Artificial Intelligence in Urban Farming, and tips for indoor urban farming beginners.
Most importantly, Motorleaf is now live on Kickstarter! If you find this article interesting, check out their kickstarter page!
- The Role of Artificial Intelligence (AI) in Urban Farming
- The typical scale of indoor farmers in acreage
- How the components of the Motorleaf system help indoor urban farmers
- Tips to Improve Indoor Growing: Conclusions from the Motorleaf Data
- How to design a "blueprint" for your urban farming system
- Rapid Fire Questions
UV:Can you talk about, for people who don't know the background of Motorleaf and how it started, the cofounders, kind of what the founding principles were, that kind of thing?
Ally: Well the background to how Motorleaf started really starts with my co-founder Ramen Dutta, and he had an education at McGill university in agricultural engineering, so he always had an interest and a passion for agriculture, and he's been growing indoors since he was a kid basically.
He left university and for the better part of 15 years he was at an IT company, he was still growing indoors, but one of the recurring problems whenever he wanted to go away (from his indoor growing operation), he'd either have to find somebody who really knew his system to look after it while he was gone or he had to basically expect is crops to do really badly while he was away or die.
So that had a pretty big impact, because if he would go on vacation for as long as he wanted, all of his crops would die.
So Ramen went online and thought there should be some sort of solution, like one on a smart phone, like Google, Nest, for example, something like that, but for his (indoor growing) system, but he couldn't find one.
So he basically started hacking together the first version of Motorleaf back in 2015, and shortly thereafter, we started working in the same co-working space, and started showing it to some other people, and they said "Wow, build me one!", and at that point we decided to try to bring it to market, and our ambition wasn't particularly grand, it was, "can we bring it to market and make enough money to just sort of supplement our income" was really as much as we wanted to do.
Then we went to the Indoor AgCon conference in Las Vegas, in late 2015, and we spoke to Mike Betts, who is the director of investment for AgFunder, and he said he liked what we were building and encouraged us to not just try to reach market but start a proper business and get going.
Then next, by luck almost, we applied to the Founder Fuel accelerator program in Montreal, and unbeknownst to us, the theme of the cohort that they were starting was Artificial Intelligence, and they were really one of the first groups of people to start speaking to us, really it was the first moment when a 3rd party looked at what we were building and really started to look into the size of the market, and said "Hey guys, what you're building has bigger applications than what you were first thinking about", and we finished founder fuel in July of 2016.
From that point forward we decided we would really need some funding to accelerate the business, and that's what we've been doing since then.
UV: It's just you two co-founders right?
Ally: Yep, Ramen and myself are the co-founders and and including some of the part time people that work with us we're at 10 or 12 people right now.
UV: Why does agriculture need AI? Are there more simple solutions that can get most of the problem solved? What's the key difference here?
There's a few really key differences.
When Ramen was looking online for a system he really wanted something that was wireless and scalable so that he could access his data, looking at his growth.
He wanted a true online and offline system that could talk to the AI side of things, even if you set things on a timer, or as an example, you set things to keep your pH at a certain level, that doesn't take into any consideration data that your plants might be emitting that you could capture that could tell you that they actually need something to change immediately.
Maybe that data will tell you, based on all kinds of different variants, if you set or change something now, something bad will happen if you continue as you are.
Really, In a way, one of the definitions of AI is that it's able to do something that no human can do manually, so a very simple example would be that with our system, every four seconds you're testing your pH level.
Now, you could say that a human could do that every four seconds, do that test every four seconds, but through machine learning, could a human predict a very complex algorithm based on thousands and thousands of data points every single day?
Probably not. People need indoor agriculture to start getting smart because it's not a place with (this type of) technology, we're giving them shortcuts and answers that otherwise would take a hell of a long time to figure out.
UV: For the people you're working with the actual customers, can you talk about their scale? For someone who is just starting out and thinking about starting a system like this themselves, what are the square meters or square feet that their working with and what is the range there?
Ally: One of the best things that Ramen did in the early days is that we had this one prototype that was the first big unit, it had everything, it was the first prototype, and it was very expensive to produce.
But he, by design, opted to include all the components, and by connecting them wirelessly by the radio, it enabled you to add on more and more (components), it enabled anyone to grow and use our system from the size of a closet up to 5 acres, and that's simply using one, main control, which is the "heart" unit, and from there you can add on other pieces and start monitoring and controlling everything else that you might need.
UV: Is that as simple as sticking a sensor in the soil or growing medium? What does the hardware look like?
Ally: The four units that we currently have on the market (The "Droplet", "Driplet" , "Powerleaf", and "Heart") (we have more coming out soon), cover the different aspects.
If you want to break it down into 2 categories, they are:
1) monitoring, and
So on the monitoring side of things, the "heart" has the hard drive inside of it, it allows for the alerts and communicators with the web tool, you can have a view of your indoor growing, you can also monitor motion for security, and every 4 seconds it's monitoring your air temperature, your humidity, and things like that, so that's the first unit.
The second unit is called the droplet, and every 3 or 4 seconds it is monitoring your water level, water temperature, your pH level, and your nutrient level, so that's four different things being monitored every four seconds near your reservoir tank (compatible with both soil and hydroponic style growing).
The next unit we have is the driplet. The driplet can deliver pH up, pH down, nutrient A, nutrient B, depending what the grower has determined to be the optimal one, the optimal application of nutrients that it needs. '
UV: I'm assuming the load the nutrients physically at the growing site? They can add the actual physical nutrient or buffer or whatever they're doing right?
Ally: Yes the driplet basically sucks up, pH up pH down, Nutrient A, Nutrient B, from whatever containers they want to use and then it pumps it into the reservoir tank and that allows the readings that the droplet gets.
The last unit does a lot of things, and it is the "Power Leaf", and with the Powerleaf you can connect any 2 pieces of crop equipment, as an example, a humidifier or a de-humidifier, but based on what the sensors from either the heart or the droplet are telling you, you can turn things on or off automatically, so for example, using data from the water reservoir tank could be "hey my tank is down to the last 20%, I'm going to have to turn on my reservoir".
UV: You can customize solutions using the (4th) Powerleaf component, and make your own custom whatever it is that you want to do?
Ally: Exactly, you might just want to buy one powerleaf you may want to buy 100 of them.
UV: Getting to the conclusions that you can draw from getting access to this data, based off of the data that you're collecting, if you had to summarize a couple ways that people using your system or people not using your system to really quickly improve how they're growing, what would those, say, three things be?
Ally: People don't know, and perhaps don't even record how much time they spend manually doing anything, if you're talking about a commercial grower, they're running a business.
So there's the growing aspect and the running the business aspect, so what we've found is that you can 70% of the time that someone is likely being paid by using our system. That includes adjusting the pH and all the rest of it.
The data is allowing people to understand what has happened, during their grow. So we also, have time lapse video, you see your whole grow as like a movie, but instead of seeing just eventually what happened, you can see, on this day, this happened, on that day what happened that made my plants droop, my pH suddenly collapsed, did my lights not come on properly?
Imagine being able to go back in time and look in detail at everything that happened during that grow that's a really useful tool.
Not only just looking back, imagine if you had a fantastic grow that you want to replicate, look back, see exactly what your settings were, then replicate that for the next grow.
UV: So to make sure I understand correctly, one of the things that you've noticed or one of the mistakes you think people make is not properly accounting for the time they spend manually working and also they don't pay enough attention to past results that have been good or bad, and trying to avoid or replicate those results.
Ally: Yes. Also just the manner in which they record those good results. How many people do you know who literally write things down on scraps of paper and then say "I just lost it that was three months ago where did I put it?".
Imagine having one place where all of your data is clean. It's data that is very accurate as well it's not like "oh I had a quick look at the temperature gauge it looked around 20 degrees Celsius" but now you actually have hard data that is indisputable.
UV: A lot of people ask us about a word, it comes up a lot, that is "blueprint". The essential components for a system, can you talk about the core components and what a blueprint might look like (for indoor urban farming)?
Ally: We've designed everything to be system agnostic, anybody that's using anything in any manner can use us.
But what we've found that comes up again and again when we're being asked, "Is it going to work with X", "Is it going to work with Y", is are they soil based or are they using a hydroponic type of system or aeroponics.
The most important thing I think is are people using a water reservoir tank with the pH and nutrients directly (hydroponic) versus someone who is (growing in soil).
Lighting is also incredibly important, we're talking with many manufacturers who are very interested in knowing how effective their lights are and capturing data while the grow is happening.
Having enough light and the right kind of light is really important more and more for people.
The grow medium, we know that people want to test, as an example clay pellets for one grow and coco for the next one, so understanding what is the grow medium for the crop you're using.
Everything is a recipe. If you're making a cake (for example), your ingredients (can be different), using one type of flour vs another type of flour, tastes completely different.
Some people need to figure out, "What is my ideal set of ingredients for my set up", and then stick to it if they're happy with it.
That's where comparing the data really helps, because if you have a record of exactly what you're doing and have been doing you know what works and what doesn't work and what exactly changed, and know, "what we're the things that worked?".
UV: Rapid fire questions: What's your favorite fruit or vegetable?
Ally: Strawberries because they should be grown in every land and we hope to do that. Favorite vegetable would probably be...probably a carrot.
UV: What's the most common crop you see your customers growing?
Ally: We don't ask, based off the information that we do have, I would say that it is probably a split between micro greens and leafy greens.
UV: What's a good book that you've read, about agriculture or urban farming that you would suggest to somebody who is just starting out? It doesn't have to be about urban Ag it can be any kind of book.
Ally: I would say the one that resonates with me, because everyone in agriculture has to be patient and philosophical about things, is a book of quotes from Winnie the Pooh. Winnie the Pooh has a good outlook on life.
UV: What's the biggest or moist surprising things you've learned in your Motorleaf journey?
Ally: How hard it is to build hardware, it is not a joke. It's like you have to start 2 different companies at the same time. Most people don't appreciate how big of a challenge it is to get hardware to market.
UV: How many times harder is the hardware than the software?
Ally: I would say it's probably 10x as hard. If we start talking about artificial intelligence and making that work, which is a whole other part, that is 20x harder.
UV: Last one I have is what can you tell us about your launch, what are the details, what's the website people go to, what's the pricing, any other relevant information?
Ally: We've been in beta since the summer of 2016, publically launch on Kickstarter Feb 21st, and the best place to find that information is Motorleaf.com, you'll find links to the Kickstarter on that page, also if you sign up for our newsletter on our website, under the contact box area, you will automatically get emailed the early bird special!
UV: What is your policy on data ownership?
Ally: I want to be clear that people don't have to share data with us, there's always a question around who owns the data, so we have an offline and an online mode that people can use that will (provide the option of private data).