Protocols and standards, towards an openness revolution in manufacturing
alternate title : if you’re nerdy and you know it clap your hands
Hi. My name is Renan, I’m the CEO of OPEO Startup Studio, a startup studio based in France and specialized in manufacturing. Alongside incredible founders and world-class collaborators distributed across four organizations working closely with sixty clients, we intend to help manufacturing companies become sexy again, create jobs in underserved territories, and unlock value using the best methodologies out there along with the best tech can offer, for manufacturing.
I’d like us to have a chat about protocols and standards. You may find it interesting because :
- While protocols and standards may be one of the less sexy and nerdiest topics ever, it is widely recognized to have been one of the factors behind the tech revolution, the commerce revolution, and in our opinion is one of the top factors holding manufacturing back ;
- This opinion is fueled by the opinions & experiences of a team of sixteen people uniquely positioned, having a technological background and working daily in factories, pushing digital and technological products ;
- There are jokes all around the article, pretty pictures, and even a wedding photo with containers.
On the importance of standards, protocols and openness
“Standard” may be a dirty word in your book. In some heads it rings with rigidity, mediocrity, and getting people to act robot-like. Truth is, standards are a large part of what the world is built upon.
Just a lil’ bit o’ history : from the amphora to the shipping container, the physical world leads the way …
One of the first ever true standard was the amphora. Amphora was a nice invention made somewhere around 3000 B.C that allowed people to keep their goods, mainly agricultural ones, from getting stale.
While the original intent was conservation only, it literally jumpstarted commerce because :
- Transportation of goods was now possible and standardized, putting a transparent price on transportation per amphora (predictability) ;
- There was a mutual agreement on the size of amphoras (actually, this came a little later, but that’s beside the point) so everyone could start working without a complex informational workflow (openness) ;
- There was a mutual agreement on the rough quantity of a given good was due per amphora, and amphora were re-usable(quant ability/re-usability) ;
- The amphora could be transported irrespectively of their content and transported together. More importantly, one could optimize the inside load of the amphora and let someone else take care of the global amphora-to-amphora optimization (strong interface, no dependency) ;
- The amphora could hold almost any goods, and get transported by any chariot (interoperability).
So, by accident, the amphora became one of the widest standards in use and benefited the whole commerce ecosystem.
Almost same cause, almost same effects. Around 5000 year later, the shipping container was created.
In the early 30’s, a supply chain specialist reportedly fumed at the inefficiency of the load on his boats. He took on a mission to improve that efficiency and invented the shipping container. All hail McLean (yes, it’s his actual name).
Much has been written on the shipping container revolution but one can immediately understand the benefits, along the framework we used :
- Predictability ;
- Openness ;
- Quant ability ;
- Strong interface, no dependency ;
The shipping container has been recognized as a driving force in global trade, bringing transportation drastically down, and other factors in the explosion of global trade.
Those two examples underline the critical importance of open standards to efficiently optimize complex webs of people working together.
So, the physical world is full of such standards, such as palettes. What about the digital world ?
Just a lil’ bit o’ history : Tech takes the lead, from TCP-IP to APIs through git & docker
One of the founding bricks (there are others, notably hardware) of the digital revolution was the TCP-IP protocol. Its invention, basically, stand on the premise that every bit of information trying to find its way in the network should be treated as equal, based on weight. You may have heard about it with the “net neutrality” wording. It jumpstarted the internet because that protocol ensured :
- that anybody could participate to the network (openness) ;
- that any form of information would be treated equally (fairness) ;
- that the standard was here to stay (predictability).
As one can see, those characteristics are aligned with the standards, its counterparts. The web was launched, and it would change history.
Some twenty years after that, a guy named Linus Torvald (yes, he inventend Linux) grew dissatisfied with the way developpers were working together on code. He found it messy, not open enough and not clear enough. So he went on a mission and invented the git methodology, which can also be described as a protocol.
Without getting too technical, git is a protocol enabling people to collaborate around any source of information. Its main characteristics are :
- Anybody can go and propose to change or add to the information, with certain rules to accept it as the new single truth (“main branch”) or decide to create an alternate truth, concurrent to the existing one (“fork”). Yep, that’s what we could call openness ;
- There are no gatekeepers to re-using open projects, which made code collaborative and wildly re-usable. Nobody re-builds every bit of code, and an estimated 95% of lines of code of any given project is re-used exactly as is from other repositories, which can be described as re-usability.
Long story short, this new protocol quickly imposed itself as the standard way to code. The most contributed repository in the history of github, Linux, has 724,299 contributions for around 100,000 contributors. And none of them got paid, and there was no formal project manager per se. And it’s the most wildly used operating system.
After the git revolution came the API revolution. An API, or Application Programming Interface, is a way for two different programs to send eachother informations over the web. Typically, the client sends information, the server process it and gives back another information. API specify the type of information the server can receive and what it will do, along with other type of informations. API share a few characteristics, that I’m sure you’ll be tired of reading :
- Guarantee the behaviour of server and client (format of the information coming in and out), that’s transparency and predictability ;
- Recognizing that there’s no particular need for a gatekeeper preventing anyone from using the API, that’s openness.
In a nutshell, protocols, both humans and purely software-based, started to rule the world several years ago thanks to a radically new way of considering value creation and openness. As a result, code tend to be written only one time and then re-used through the git protocol, propagated through thanks to an open protocol, and different parts of the ecosystem are interacting without human interaction or any kind of gatekeeper preventing the whole system from improving itself. It is one of the key components of exponential scaling : re-usability of existing assets. Colloquial term is “compound interest”.
And now for the bad part : the state of standards in digital manufacturing
Oh noes !
Any passerby can attest : current state of digital affairs in manufacturing is not great. In addition to legacy inflexible systems, interoperability is probably at a high-time low.
Out of nearly twenty companies surveyed for interoperability of their manufacturing programming for machining, exactly zero had any kind of interoperability. Let us say it another way : we estimate at more than 60% the number of lines of code or pseudo-code that were done in those companies were done at least twice if you count their other factories, suppliers and clients.
Simply put : standards are not interoperable, not open, and not transparent.
There is some light, notably in the car industry re-using modules. But those modules are closed-source and more a case of coopetition than real cooperation on open source.
Translated to the tech world, it is roughly the same as if developers had to rewrite their programs from scratch without using libraries. No way tech would scale. Well, that’s one of the factors explaining why manufacturing is prevented from scaling.
It’s the economy, stupid !
There are many economical cases to be made about this state of affairs. Mainly,
- Actors who could enforce standards are few and far between (basically, big manufacturers only) , so it is in their best interest to lock down part of their ecosystem, thus enforcing a local optimum rather than a global one (what we call the game theory argument) ;
- Codesharing was simply not possible, and open sourcing was seen as a hobby, when manufacturing started mechanization, and so the expensive systems put in place are not flexible enough to let that revolution happen (what we call the innovator’s dilemna argument).
Those arguments make sense. However, we also think there is largely a cultural explanation.
It’s about the culture also
Historically, Silicon Valley started with a culture of openness and sharing. The open source revolution started here for a reason.
Silicon Valley started famously with a bunch of nerds, weirdos and definitely not the suit-and-tie type. Their new way of “doing business”, which involved sharing to the extreme, transparence and a new brand of coopetition ultimately defined the modern world in ways few understand. And other sectors, such as manufacturing, are still closed.
Manufacturing is, by nature, closed loop on its ecosystem. Unfortunately, manufacturing companies are keen on protecting their trade secrets. So collaborating on open standards, open libraries of code is not the standard — far from it.
It’s hurting everyone
This state of affairs increase costs, reduce velocity, and prevents a boom of the industry.
It also prevents manufacturing from being able to recruit extremely high potential, who would rather work on very big impact projects re-using existing blocks, rather than wasting time creating the same blocks over and over again.
Cannot we do better ?
Let’s go !
There is a blueprint for what manufacturing could do : there are troves of fundations aiming at creating open standards in technology. The largest and most renowned one, of course, is the Linux fundation. It enabled the largest software development in history.