OPTIMIZING NATURE
Many of the previous journal entries were devoted to explaining how our system works, how we got started, and how we plan on using our technology to rid this world of toxic waste and the supply chains that support them. Now, I would like to take some time and explain how we make these systems run as efficiently as possible in any environment.
First, let me introduce myself. My name is Rob Auchincloss and I am the Chief Technology Officer at EDEN. I studied aerospace engineering at the mountainous Arizona campus of Embry-Riddle. Since college I have had an incredible journey working with different designers, architects, engineers, scientists, and founders on a variety of amazing (and some ultimately not amazing) projects.
While at ERAU, all engineering students were required to take the course AE 430: Controls. Now you’re probably imagining what I did whilst signing up for the class - learning about how controllers and other related systems interact with aerospace elements & operations. I was surprised day one when the instructor spent the first fifteen minutes of class explaining that it would take nearly a full semester to begin to understand the complex theory involved with control systems.
A control system manages, commands, directs, or regulates the behavior of other devices or systems using control loops (I always appreciate and am thankful to Wikipedia for having the most concise base definitions) I will save you all the dross detail and break it down into its simplest form. Essentially, any component of a control system is “loop” (like shown below) that allows the system to understand and “react” to given inputs. This may seem familiar and it should, our own brains (when behaving nominally) function very similarly.
There are two types of control loops - open and closed. I will explain both using a simple analogy: imagine a car’s cruise control. Now, up until about a decade ago, most cruise control systems worked quite simply - you would set a speed and the car would maintain that speed until you either turned off cruise control, braked, or crashed. This is what is referred to as an open-loop control, or a loop that is only taking in a single (or in some cases multiple) set point, in this case speed. But recently a new kind of cruise control has become standard - one that can slow down for traffic, obstacles, or slower cars ahead and even match that speed. This is what would be referred to as a closed-loop control, or one that is constantly reacting to outputs of a system, and optimizing to adjust future inputs. A closed-loop control system is what is used in all EDEN systems.
Now utilizing differential equations (a subset of mathematics that involves one or more derivatives of a function in its practice) we can optimize control systems by continuously optimizing the cost function - or simply put: ensuring that the minimal amount of resources are used to get to the solution. Now for most systems, an optimized control system is efficient enough. However, for EDEN, we needed something a lot more powerful.
Machine Learning (ML) is a field of study that governs algorithms that can learn from data, predict unseen or unknown data, and learn & act from it. Most of the time a company or individual claims they are working in “AI” (artificial intelligence) they are really just using ML. In our case, ML will allow us to have systems that are constantly learning and optimizing based on the variety of inputs that could be placed in our system.
Since our systems intake waste in a mixed-stream (as in the combination can be changeable, variable, and diverse), our systems need to be able to adjust on the fly and always optimize for the current system of waste. This also gives us an incredible competitive advantage, as most existing “waste to energy” or similar solutions need very specific conditions to be met in order for the process to run, let alone efficiently.
As our systems continue to process waste, the amount of available data to train our system will increase exponentially. The more data we are able to collect, the better our model will become. We collect all of our users’ data and encrypt it before sending it to our central command where it becomes part of our ever-optimizing model (we will never sell or share our users’ data. EDEN takes an incredibly strict approach to privacy).
We are excited to continue to work hard to get these systems out to the entire world. If you are someone with background knowledge in data science, mathematics, or machine learning, we’d love to have you join our team and help us change the world.
Be well,
Rob Auchincloss
