First impressions from kickstarting a new language – CircuitData
By Andreas Lydersen, CTO Elmatica
This January (2017) in Norway Elmatica, a broker of printed circuits with 45 years experience in the business, started an initiative to have the whole industry talk the same electronic language – CircuitData. The reason for it was simple: throughout the business, thousands of hours are spent every day on interpreting each other’s requests for quotes and orders. Both time and money wasted.
This attempt was not the first one to create a common language, but previous attempts have focused on creating formats for single files that would cover the whole supply chain, in effect forcing everybody to change their whole digital ecosystem. Elmaticas initiative, however, allowed partial communication to share a language, thus slowly eating into the manual processes that are so time-consuming. It also respects the fact that supply chains are complex and has a tendency to fight changes.
Open source is king
Elmatica decided to use a very successful method adopted from software development, Open Source. Today Open Source software in effects powers the Internet. This also means that anybody who contributes, big or small, in effect gives away their work. Nothing belongs to anybody – it is maintained and developed by a community of users. To organize the end result, a board put together by different players in the industry will make sure that the master branch of the language is relevant at any given time. If anyone disagrees with the board, they are free to fork out a branch of their own and continue working on that.
Anybody can make a language – many of us actually did when we were kids. We designed secret signals and sounds that only our friends would know the meaning of. It is when you go from simple to complex that it gets hard. Whereas kids easily see the need for a new language (keeping things secret), adults are less eager to adapt to that idea. Selling the idea of a language is selling something abstract, and abstracts are hard to sell. It requires the creation of a common vision, and a desire to follow a common path. The easiest path to choose is probably to boil it down to something concrete – “this is how you use the language in a piece of software”. But then again, this might be a bad idea. By narrowing in too much, you risk losing out on all the ideas that could come out of a crowd of people sharing the same vision.
In our case, we see both. Some people need actual products to sell or buy, others sell inspiration and accept a common vision. Both strategies boil down to similar questions like:
- When can we expect to see the first adoption of the language in a format we can use?
- What can I/we do to help?
- When will it be finished?
- How many are using or contributing to this?
No common language – no automation
In my experience, selling an abstract idea works best if the people receiving it have a tuned mindset to accept a common vision. However, projects still need to be somehow tangible, since abstract ideas have short lifespans if left abstract.
What generally is accepted though, is the fact that the lack of a common language creates manual labor. Most companies working in the PCB industry have ongoing projects to automate production or other repetitive tasks. Still, anything that has to do with parsing incoming requests, are pretty much-unploughed ground.
On a visit to China recently, we had discussions with several large manufacturers. What we learned was that departments that handle incoming requests, are often one of their largest departments and also, except management, their most expensive. A simple analysis of their tasks shows that if they received information in a digitally parsable and common language, and this language had the possibility to show conflicts throughout multiple layers, they could reduce the number of staff with 50%, or achieve a 100% increase in productivity.
Any project with this kind of potential return on investment is embraced by the management. However, since this is an abstract idea rather than some piece of equipment they can buy and install, the questions are pretty much the same as the ones stated above.
Feed it, grow it, develop it
So how do you move a whole industry from “What can I buy?” and “When can I buy it?”, to “What can I do to help?”.
Actually, I’m not sure you can. The fact is, it needs to catch on like a bushfire. Let me give you an analogy: The language you are speaking every day is not specific – it is abstract. It is built up by components as grammar and words. Diversified by dialects and slang. Words are added and removed by people like you and me, often the youngsters, not by a central committee (at least not at first). The vision of your language is to unite you in common understanding based on sounds and signs. When the need to simplify or communicate something new rises, we create new words or even change the grammar.
As this is not centrally governed something peculiar happens: Two or more people can adapt the language to cover their needs to communicate, and it immediately works. As long as any two parties share a common understanding they gain results.
CircuitData has the same potential. Any two players in a supply chain will gain immediate results from talking together without interpreters. That is how it will catch on. It will replace manual routines one by one – not all at one time.
The idea of CircuitData was born one year ago, still, its journey has just begun. We encourage all to participate since a language needs to evolve and constantly develop. A lot has happened since we started this journey, stills it’s only the beginning of the adventure.
Previous article about CircuitData.