Since launching the company in 2014, we’ve created over thirty prototypes. That’s an average of one prototype every three weeks for two and a half years! Some prototypes were better than others, and some were definitely dead ends.
A question many people ask is how did we go from this jumbled mess:
to this lean mean fighting machine?
We’re going to tackle this question in two parts:
- Our Process for Hardware Prototyping
- The 3 Types of Prototypes We’ve Made
1. Our Process for Hardware Prototyping
From day one we’ve had a focus on rapid prototyping.
We define rapid prototyping as building, testing, and iterating our product as swift as possible, all the while continuously getting feedback from our earliest supporters. This often meant our core assumptions about our product were challenged. The feedback was invaluable and led to us changing parts of the product such as the user interface of the app or reducing the legs from six to four.
In the beginning, we hand-molded and 3D printed parts. Everything we did, we did in-house at the Bristol Robotics Laboratory, or in Silas’ student housing. At the very beginning, our goal was to show that we could build a rudimentary MekaMon, and to see if people were interested.
2. The 3 Types of Prototypes We’ve Made
There have been three main types of prototypes that we’ve created:
We define an aesthetic prototype as one that exists to test out the look of the robot. Often, we will do this in various materials or colors just to see how it looks and feels in the real world.
These are photos of some of our early looks prototypes made out of Styrofoam. You can see the influence of the earliest designs in our latest looks prototypes. Sadly, the shuttlecock base was deemed “unnecessary and over engineered” by Silas, only to be replaced by four legs.
With hardware, the aesthetics are often restricted by the desired function of the prototype. For example, moving the motor and PCB by 4mm in one direction could drastically affect the aesthetics which leads us to…
Functional prototypes serve to demonstrate the product, and to test out hardware, materials, and electronics configurations. With functional prototypes, we want to build them fast and then test. Quick iteration is key here.
Check out this GIF of one of the mid-stage bodies going through the walking motion.
Or when we tested out a shooting animation:
A few months ago, we had to alter the aesthetics of the leg to accommodate a new electronics layout. We had to move a motor by a few millimeters and it would no longer fit in the old leg design.
To create a raised aesthetic feature where we needed to the motor to be placed. That allowed us to kill two birds with one stone. The MekaMon gained better functionality and a more detailed design.
Notice that in images 1-4, there are cables running into and out of the thighs? From the very beginning, we knew that the final version of MekaMon would be required to have no cabling between the hip and the leg. In the final design, the entire leg will be modular. There will no longer be a cable at a connection point. However, for the sake of rapid prototyping, we kept the cabling as is.
It wouldn’t have made sense to spend time and resources on figuring out how to make the entire leg modular while there were still so many changes to be made.
Follow us on Twitter or Facebook to see how our prototypes have evolved. Every other Monday, we feature a specific functional prototype and what we learned from it. We look back at our past prototypes and relive the not-so-glory days.
While the hardware team rapidly churned out prototypes, our Head of Software, Jonathan, needed something to begin testing code on
This leads us to…
We define developer prototypes as “as done as possible at this stage in development.” These would be prototypes that are as functional as possible that could be used for demos and to test out the gaming software on.
An example of this: Jonathan might create a prototype to in order to test out specific software applications. These developer prototypes don’t need to be a fully functioning robot, but rather, just a subset of the hardware.
If you live in Cardiff and frequent coffee shops, you may have seen Jonathan waving one of his augmented reality head prototypes in front of his computer like a loon. We swear he’s normal (well, almost normal). He’s doing this so that we can figure out the most functional shape, color, and structure of head of the robot, in order to implement augmented reality.
In the coming weeks, we will also be detailing our software development process. Stay tuned ;).
The post AND YOU GET A PROTOTYPE! (Or The Process Behind Creating a Prototype Every 3 weeks for 2.5 Years) appeared first on Reach Robotics.