Sanne 'SpuQ' Santens
The Internet of Machines
A whole segment of digital solutions requires a reach beyond the perimeters of the cloud; applications with a fleet of machines ('things') deployed out in the real world - ranging from devices performing simple measurements and sending telemetry to the cloud, to complex machines interacting with the world through sensors and actuators, making autonomous decisions as part of an interconnected federation. We called it the Internet of Things (IoT) - and, for lack of inspiration, we rolled with it, I guess. Though I like to call it what my system software teacher called it - still not sure whether it was a slip-up, or very deliberate, maybe in reference to The Matrix: the Internet of Machines.
The Real World
I like to describe myself as an inventor with a passion for intelligent machines and a fundamental desire to explore ideas through experimentation - though I've learned not to bet the whole farm on my next brilliant impulse. In the past I had to abandon countless ideas because the project's technical requirements moved too far beyond what I could confidently take on, budget became too tight, or - within those constraints - ability to deliver a prototype that could actually withstand the testing conditions of its real-world deployment long enough to deliver representative results while allowing small changes and gradual improvements to evolve it into a viable solution. When your solution can operate within the limits of the datacenter you spin up a VPS and start iterating on your idea with very little risk and watch it unfold into success or failure - fast, safely and cheaply. When your idea needs access to a physical environment for measuring its conditions or manipulating its state, you are required to deploy tailor-made hardware exposed to the unbending laws of physics and unpredictable hick-ups of the universe - bringing a whole different price tag, skillset and uncertainty into your experiment.
If you're the type that's tempted to hot-glue some breakouts to a Raspberry Pi, so you can have a decent approximation of your application-specific hardware interfaced with your favorite high-level programming language and extensive frameworks, quietly praying you got it right first try while you're casting the whole thing into resin for weatherproofing, in order to give testing out the concept of your idea a shot? I see you - I've been you. I started to take my ideas more seriously, and developed the Edgeberry project to support me into putting this into practice.
Raspberry Pi Equipped For the IoT Edge
I come from an Embedded Systems background, because I always loved creating physical things. For imposing your will onto your electronics, you need 'compute' and software as ideas get more complex - eventually concluding that a Linux server with a microcontroller's form factor would be great. Lucky that I reached this point in the early 2010s, just as the first Raspberry Pi was released. Being able to combine high-level software skills and a wide range of (experimental) frameworks with low-level 'embedded' hardware feels like having super powers. Though the Raspberry Pi of itself is not well suited for leaving the safety of the lab environment. From my years of experience in prototyping IoT applications, and abusing countless Raspberry Pis in the process by throwing them out there anyway, I created the Edgeberry ecosystem - actually equipping Raspberry Pi for the IoT edge, and making it look so dang good that your manager (or customer) will love it before they even know what it is.
The Physical Pillar
I came to think of the Edgeberry project as the bridge interfacing the real world with the digital world; with a virtual endpoint in the cloud and physical endpoints deployed in the real world, shaped like things - allowing my solutions to seamlessly span the whole landscape. The Edgeberry Zero is its column in the real world; generic infrastructure where my application builds upon...
Read more »
