If your main source of information so far is Caleb’s old TED talk video from December 2015, I recommend that you read through the forum to familiarize yourself with how the OpenAg project has unfolded in the last three and a half years.
The PFC design featured in the TED talk is no longer supported (see Update from OpenAg - Please Read), and OpenAg recently recommended that people do not build their current PFC design (see this entire thread: Meet "PFC_EDU," the new PFC v3.0). It sounds like Caleb’s team is working on a new design that they intend to have ready for the MIT Media Lab fall 2019 member meeting (probably late October; see comment #15 of PFC 3.0 BOM not complete? (according to Eagle file))
Also, Caleb’s twitter feed (@calebgrowsfood) has links to lots of his more recent interviews and conference talks. In particular, I recommend that you look at his stuff from mid-2018 to the present in order to get a sense of where the OpenAg project is currently headed.
The community designed MVP food computer is fairly popular, currently supported, and suitable for growing small quantities of food (see $300 Food Computer - MVP) . Also, you might take a look at the stuff @house is working on in Korea (see OpenAG in South Korea and FoodCity in FAN5(FabLab Asia Network 5)) and the stuff @VictorB is working on in Spain (see Food Computer - Gipuzkoa style).
[edit: Oh, and I forgot about FarmBot and AstroPlant. They are separate projects from OpenAg, but many people here are interested in them. It’s possible that FarmBot might be a better fit for NCAT compared to building a food computer, assuming your goal is to grow food to feed people rather than doing phenotype research (food computer yield / $ is low).]
[edit #2: Also, the Bright Agrotech youtube channel has lots of information on starting small indoor farms using aquaponics, hydroponics, and vertical growing in greenhouses. From the sustainability angle, you might like their focus on how to produce food cost effectively in small indoor farms.]