If one simply abstracts all issues to their fundamental quantities of mass, density, productivity, photo-efficiency, inputs and outputs as a mass energy balance the challenges can be seen as applicable to all potential solutions. Just dealing with all of this mass energy movement is a significant challenge regardless of what novel or natural configuration is implemented. These are the fundamental challenges to economic competitive viability inherent to all solutions here.
As far as aquaponics (and hydroponics) that is something I designed our systems for cross over application as a second phase. I have automated my own hydroponics in the past and learned key lessons about reliability for such environments which I have now addressed in our lab systems. I also developed the software architecture in such a way as to have all channels, signals, components, sub-systems, systems and platforms within an enterprise to be dynamically deployable/configurable/removable for integrators to pick from pre-built software/hardware plugins. Our automated ponics systems will have interconnectivity for remote monitoring on operators via smart phones or internet. Our plan is to offer everything from components to complete systems and everything in between.
As far as sensors: EC, ammonia, nitrites and nitrates
My company has EC and ammonia as well but its not cheap at this point because we don't have volume. You are going to have challenges with cross-talking between sensors. As far as low cost solutions if you solve the problems yourself at the circuit level then you can produce interface circuits relatively cheap but you need to understand the problem. I am working to capitalize on this so I will not give the solutions we developed but the information on solutions is out there from other sources.
In hydro you use TDS targets for nutrients which is based on EC but that's also because you know what you put in where aquaponics is making it. I have not addressed the aquaponics side at this point in R&D but I think the highest probability for success is probably in finding good relations between EC pH and Ammonia for estimating nitrate/ nitrite levels from empirical studies with nitrate sensors initially recorded manually then incorporated in software as cal curves/surfaces or lookup tables. Here is a nitrate sensor but I have never used this product and it still does not address multisensing platform issues:
If I were you I would do extensive experiments with your sensors first before any automation. You are going to find issues. That's where I think we have an edge because I learned these issues years ago with my hydroponics efforts and I invented all kinds of solutions over the years simplifying things down to mass producible products targeted for low cost implementation. This is not easy to do for low cost but eventually you will see what we have made penetrate these markets and you will realize why we did what we did and I believe it will likely take a center stage for the sensing system of choice for automation of all types of these measurements.
The most powerful Law of Nature is Time. It is finite and we all will run out of it. Use this Law to your advantage, for it offers you infinite possibilities...
Want DP delivered to your inbox daily? Subscribe here: