It all began with the first industrial revolution. With the transition to new manufacturing processes and steam-powered equipment, the world saw the birth of a ‘culture of labor-saving and efficiency.’ Soon after, the technology innovations paved the way for mass production of specialized goods using electric power(the second industrial revolution), eventually leading to manufacturing automation with electronics and IT (the third phase). To satisfy the ever-growing demand for manufactured goods, the use of energy has been escalating unchecked ever since. Even today, in the era of Industry 4.0, the continued exploitation of both renewable as well as non-renewable energy sources, coupled with the use of inefficient technology, raises serious concerns: is the manufacturing world striding toward an unsustainable future?
Stirred by these predicaments, five years ago, when Vincent Sciandra completed his Ph.D. in Computer Science, he envisioned creating a manufacturing ecosystem where energy is always consumed in the most sustainable and efficient way. He wanted to help design a factory where the machines are optimally operating within designated limits, facilities are efficiently reducing their carbon footprint, and floor managers are constantly seeking new pathways to implement a more energy-efficient and reliable production practice. In an age where manufacturing activities account for around 54 percent of the planet’s total energy usage, M. Sciandra’s aspiration seemed nothing more than a far-fetched vision. However, the headstrong M. Sciandra did not falter; he wanted to dive into the core of the problem. At that juncture, he realized the extent of incoherence between machines and software in the industry; “The communication between the machines was very hard to establish,” says M. Sciandra.
This is where M. Sciandra was able to deduce the equation to solve the problem of energy over- utilization. He believed that if the machines interact with each other, the vast data that it would generate could help obtain insights to improve energy efficiency. Exhibiting the zeal to turn this scheme into reality led to the genesis of METRON. Founded in 2013, this French technology company leverages the clout of AI and big data to usher a new wave of innovation in the energy sector. Offering a game-changing energy intelligence platform—delivered through a SaaS model—METRON’s uniqueness lies in aggregating and analyzing energy and production data of an industrial plant in real time to identify new sources of energy optimizations. Further elaborating on how METRON is best fitted to overcome the energy utilization challenges in the manufacturing segment, M. Sciandra, founder and CEO of the company, notes, “While there are different kinds of energy- saving products and equipment available today, these alone are not enough to solve the ever-increasing energy burden.” Most manufacturing companies lack a dedicated team that constantly analyzes the flow of energy in their facilities and produces actionable insights to improve energy usage. “This is where our expertise lies,” he asserts. METRON’s energy intelligence platform leverages large volumes of energy data to predict consumption patterns and foster an eco-friendly production environment.
Unlocking Energy Effciency with Context- Aware AI
One of the most distinctive facets of METRON’s platform is its “artificial intelligence meets human intelligence” modus operandi. Explaining this strategy, M. Sciandra states, “The AI-driven platform first collects data from a facility’s internal and external sources to get insights into the operations.”The platform consists of a digital model of a factory, a sort of a digital twin, which can simulate its production and consumption processes. Then, METRON’s team of experienced data engineers and energy experts helps companies transform those insights into real savings by devising an actionable energy management strategy. Complementing the human intelligence, METRON also offers their AI capability through a completely automated “energy virtual assistant,” METRON- EVA®, which communicates directly with the operations manager of a facility or METRON’s energy experts to provide deep insights that help the managers make informed decisions and eventually cut unnecessary energy consumption.
That said, METRON has also made a striking technological breakthrough in the design of their AI engine. “We call it industry context-aware AI,” says M. Sciandra. While the first important aspect of this unique AI architecture is building sophisticated machine learning algorithms, its second pillar is the implementation of data ontology—a demonstration of linkages and relations between data. “These two, combined, strengthen our AI engine to seamlessly connect disparate data and provide more refined insights,” he explains. To further explain this approach, M. Sciandra cites an instance of a milk production factory. There are different kinds of apparatuses used in a dairy in every stage of the production. To pasteurize milk, for example, the facility uses a steam boiler, and similar types of boilers are used in a chemical plant as well. “While there might be no similarities in the usage of the boiler in these diverse realms, if we gather additional data on the equipment from different resources, it can help us gain a clearer understanding of how to optimize its usage,” reveals M. Sciandra. “On one side you have predictions based on real-life historical data, and on the other, you have forecasts based on the usage of machines in varied industries.”
The Path to Energy-Transparent Factories
When it comes to implementing the avant-garde energy intelligence solution in a factory, METRON has a two-step deployment process. “First, we have a small piece of hardware that is deployed on site, which is used to gather information from the machines directly or other energy monitoring systems that the firm might have in place,” explains M. Sciandra. The second step is the configuration of the energy intelligence platform. The time taken to complete this phase depends mostly on the size of the factory, the complexity of the facility’s operational workflow, and the ease at which its historical data can be coalesced with real-time data. “Taking all the scenarios into consideration, the implementation process can be completed within two weeks. However, in case the manufacturing facility has no past data records, we usually need three months to collect the information and train our machine learning algorithms, so that the platform can start interpreting the energy usage patterns,” expounds the CEO. Once the system is up and running, the factory can start regulating the energy consumption of the machines and maximize its output.
Elucidating a case in point, M. Sciandra mentions how they implemented their solution for a glass manufacturing facility. The facility was using glass furnaces that were designed using Siemens’ automation systems to automate some of the manual parameters of the manufacturing process. “When we came onboard, we had to map each of those parameters (whether manual or automated) to understand the constraints and how they impact the furnace’s energy consumption,” mentions M. Sciandra. Once the challenges were clear, METRON’s energy intelligence platform was able to configure the parameters to the most optimal level and help the company maximize their glass production. “With the help of our solution, the client was able to save an average of 5-7 percent of their energy on a daily basis, which accounted for annual savings of about a quarter of a million dollars,” affirms M. Sciandra.
In addition to helping the glass manufacture save big bucks, METRON’s solution also helped the company bridge the knowledge gap of furnace operators. Further describing this, M. Sciandra states, “Imagine if there are two operators, one with 20 years of industry experience and the other with 5 years, both might have different approaches to tackle a particular problem while operating a glass furnace—there’s a gap.” The algorithm that METRON brings to the table has the ability to negate this concern. “Our modules help the glass factory to operate the furnace in a homogeneous way, without being dependent on the furnace operator’s industry experience. This adds more value to our solution stack,” emphasizes M. Sciandra. Thus, even if the furnace operator changes, the company would continue to yield the same level of energy efficiency while controlling the furnace.
A Future-Ready Company
Today, this unique AI model has helped METRON earn numerous accolades from its clients spread across more than 15 countries. M. Sciandra summarizes the success mantra of their energy intelligence platform in three words: “Connect. Detect. Reduce.” METRON intends to roll out more sophisticated AI features and, at the same time, grow organically. “We recently raised about $10 million, which will be used to further enhance our AI capabilities and expand our team worldwide to support clients like Danone, a French multinational food-products corporation,” reveals M. Sciandra. Advancing into the future, METRON’s focus on creating a true energy-efficient factory coupled with its second-to-none AI engine is poised to spearhead a momentous transformation in the manufacturing realm.
See the digital issue HERE.