Project Pete
Smart Plant Pot
3d printers are wicked but they became a huge fad around 2015 with everyone having to get one. This would be fine except it felt like it all came at the expense of other more important things. For example a principle investigator in our research group decided it was more important to buy a 3d printer than a coffee machine (this was in a lab in the middle of nowhere with a crappy cafeteria which was closed outside the hours of 12pm and 2pm). Another concern for me was that 3d printers were distracting from the mastery of other equally important tools such as lathes, milling machines or even hand tools such as drills and sidecutters. Being able to make a prototype which can bear loads, withstand heat whilst also looking reasonably good is important.
​
Having said all of this, 3d printers (in particular commercial 3d printers), are phenomenally useful. It's now possible to 3d print all sorts of materials and some systems can print metals whilst approaching the strength of the original billet. I decided to jump on the bandwagon by 3d printing another engineering cliche, the self watering plant pot, the hello-world of patent attorneys.
​
The design of it was a bit of a labour of love and some of the mechanical elements I wont disclose (unless you're a major plant pot manufacturer who might be interested in technology licensing ;). So behold ... :
CAD model of plantpot full assembly (with plants I took from GrabCAD) in Solidworks and photo of the 3d printed prototype (printed in resin) with a healthy plant (which it subsequently killed through over watering)
I wanted something which was pretty funky in design and which could be easily taken apart for cleaning and so came up with a geometry which would be quite tricky to prototype in any other way than by 3d printing (honestly). I spent quite a bit of time in my sister's spare room over Christmas 2017 wiring up the circuit board inside along with all the other stuff to make it capable of watering itself and voila, it came alive. The pot runs on a little Arduino micro which fed back all sorts of sensor data to my laptop (soil moisture, PH etc). I had major ambitions to get a bluetooth module working and a little app on my phone up and running but in the end became distracted by work once the holiday season was over. The plant shown in the photo above was eventually killed by the pot which over watered it due to a dodgy moisture sensor heralding its consignment to a cupboard of other failed prototypes.
Concluding thoughts...
Whilst this project didn't turn into the product success I originally imagined, it did give me a bit of respect for the capabilities of commercial 3d printers. It was pretty expensive at the time mind you (around £150) but I figured a home 3d printer would cost more than that and there was one (instead of a coffee machine) at the lab in any case. I'm still unimpressed by domestic 3d printers and have not changed my mind about other tools still being important!
​
Another interesting area was around the algorithms that keep the plant alive. My lack of horticultural knowledge killed many plants (usually through over watering). It would be fun to explore more sophisticated approaches which can infer the plants health with a camera coupled with some image recognition software, I've got many pictures of dead plants which could train an AI watering system in the future ;). Would be interesting to hear from any horticultural experts who might have some ideas in this space...