Today, decarbonizing industry is a major challenge, for economic, geopolitical and environmental reasons. Energy renovation, use of renewable energies… what can be done to achieve this? How can manufacturers be supported today?
To help you in this transition, we have written the Dametis White Paper, in 3 chapters, so that you get the keys to guide you towards Industry 4.0!
This White Paper was written by Dametis experts:
Julian Aristizabal
Co-founder, CEO
Jérémy Barrais
Product Manager
Nicolas Duran
Co-founder CTO
Sébastien Papouin
Technical Manager
Cyril Quemeneur
Energy engineer
Chapter 1: moving towards industry 4.0 – data and human expertise at the heart of your decarbonization strategy
I. Memory and environmental intelligence
If “the development of full artificial intelligence could mark the end of the human species” (Stephen Hawking, 2014), intelligent software could mark the end of energy waste in industry. For example, the energy data involved in the manufacture of a product – from mascara to cars to bags of mashed potatoes – far exceeds the storage and processing capacity of the human brain.
An “energy memory” already outsourced
Naturally, professionals began by externalizing their “energy memory” in paper files and Excel spreadsheets. This is the case when an operator reads the meters on a printed sheet and stores it in a binder, before a colleague eventually copies the data, with varying degrees of error, onto a computer. The same phenomenon of outsourced storage occurs when various software programs slumber in a corner of the factory, storing data of all kinds, particularly energy data, without doing anything with it.
The limitations of Excel tables
Julian Aristizabal, CEO of Dametis: “Today, many industrial sites manage their energy with simple Excel spreadsheets. This involves time-consuming and hazardous data retrieval – during plant tours5 , which in the absence of automated transmission can represent 30 minutes or even an hour a day – then integration, lots of copy-pasting, rewriting errors, stacking up versions created by different users… This method takes precious time away from thinking about these data (which are in any case incomplete and unreliable compared with those that would be returned by a good EMS).”
People and A.I. together to face the ecological challenge
Software to measure environmental performance is therefore (among other things) a new factory memory. When this software is sufficiently sophisticated, the data it contains is correct, well-organized and easy to access, contextualized and non-redundant. What’s special is that this memory can be mobilized simultaneously by two types of brain: human and algorithmic. While we should be concerned that, generally speaking, in our societies “the computer comes to represent an ideal in the light of which real thought perversely ends up appearing deficient” (Matthew B. Crawford)6, we must nevertheless recognize that the human mind cannot take up the challenge of industrial energy savings alone.
Environmental software conductors in the factory of the future
Software is already essential in today’s factories as they move towards the “minimum energy achievable” (MEA), and will become even more important as production sites become increasingly automated.
“Programmable machine tools, welding and painting robots, remote-controlled forklifts, handling and assembly automata have long been part of the factory and warehouse,” recalls Charles-Édouard Bouée in his book Confucius et les automates (ed. Grasset, 2014)7
.
But the new generation of this equipment will look nothing like its predecessors, as it acquires more and more intelligence and, thanks to the Internet, will be able to connect and communicate with each other.”
In Industry 4.0, environmental performance software will be the conductor of the ecological stakes in this “new cyber-physical reality”.
The human element remains essential
Julian Aristizabal, CEO of Dametis:
“The software is based on man-machine collaboration. Because in a factory, we always end up having to deal with exceptional situations, which again require human intervention. Software doesn’t create expertise; it works on the basis of human expertise written in algorithmic form. There are also things it can’t do, like (re)calibrating sensors8 – which inevitably drift over time, in other words, the zero shifts and the data returned is false.”
II. Environmental expertise in software
Expert systems” at the service of manufacturers
A collaborative environmental transition platform should ideally concentrate the world’s human expertise in industrial environmental efficiency in the hands of each user. To achieve this, it must be a true expert system.
“An expert system is an artificial intelligence (A.I.) computer tool, designed to simulate the know-how of a specialist, in a precise and well-defined field, thanks to the exploitation of a certain amount of knowledge explicitly provided by experts in the field.”
Challenge plants against the best in the world
What’s more, the expertise provided must be rich enough to integrate the best environmental scores (energy among others) at world level, updated for all industries.
The EMS software can thus challenge utilities and processes by comparing them with the best performances worldwide. Of course, there is no such thing as a pre-packaged repository that you can simply “feed” into the software… This data must come from the accumulated field experience (plant visits and audits, implementation and monitoring of measurement plans…) of the company supplying the software, supplemented by specialized documentary research.
EMS users can enhance their energy skills
By its very nature, expert software transfers to industry the increased knowledge traditionally held by employees, service providers and consultants. However, there is no risk of losing in-house energy skills, as the software is not a “black box” from which strange recommendations and assessments emanate, but rather a transparent, intuitive tool at the service of human beings.
Generally speaking, “software must enable user autonomy – particularly in France, where there is a fairly high level of technical maturity, with highly skilled operators – by being flexible and open”. (Julian Aristizabal, CEO of Dametis).
The Dametis expert system is defined by a double transfer of expertise: on the one hand (and by definition), a process of concentrating the knowledge of our experts in the platform, and on the other, a process of disseminating this knowledge to all users (in particular on-site operators).
Our users, who are committed to using the software, are constantly increasing their level of competence and gaining autonomy in the area of environmental performance.
III. Environmental data 4.0 craftsmen
Software must first collect, second after second, a large quantity of data, wherever it may be: PLCs, ERP, MES, sensors, virtual sensors with algorithms…
And just as a cabinetmaker studies his wood before working on it – is it strong, irregular, gnarled…? -software has to “understand” its material (the data) before it can do anything with it. Is the data erroneous (sensor drift, parameterization error…) or correct?
Reflect plant life and eliminate “technical debt
“I’ve visited factories where the software data was so disconnected from reality that the tool had become unusable,” stresses Dametis CEO Julian Aristizabal. These technical debts affect, to varying degrees, 90% of the software I encounter.
What some software designers forget is that manufacturers spend their time looking for solutions to concrete problems, and therefore modifying their installations. Software has to reflect reality, and take into account the most minor changes which, put together, over the months, shape the plant.”
Examples of everyday changes to take into account
The software must constantly keep a critical eye on the data, but it must also eliminate “technical debt” by assimilating everyday changes (feeding a process, configuring a utility, etc.).
A piquage to interconnect two cold networks will change the performance of both networks, and an automation engineer can easily change an address to optimize communication between two PLCs. Intelligent software must be able to keep pace with these changes.
Would you like to obtain and read the rest of the White Paper to get all the keys to guide you towards Industry 4.0?
