SmartMelt: A Data-Driven Melting Process Optimizer

ended

Aluminium is one of the most recyclable materials on the market today. Recycling aluminium saves more 90% of the energy required versus producing new metal. Worldwide, melting aluminium scrapes amounts to around 12 billion US$ energy cost and produces ~50 million tonnes of CO2 every year. Yet, melting process is very inefficient because the operation of large furnaces is still mainly based on human decisions. This can be significantly improved by an on-line data-driven numerical simulation. Optimizing the melting efficiency by only 10% can save nearly 0.5 M€ annual cost for the operation of a medium-size furnace with a huge beneficial impact on energy saving and environment. Novamet develops SmartMelt, a unique and patented data-driven controlling system. SmartMelt solution includes a digital-twin furnace in which data coming from the furnace controller are analysed during the process and compared on-line to results coming from a very efficient numerical simulation software. A very user-friendly interface guides the operator to perform the future action at optimal times. The business model is based on annual licencing of the Software to aluminium producers, and furnace manufacturers. More than 1'500 furnaces around the world with a load capacity of 30 to 120 tonnes each are considered as a first addressable market. We expect to reach 15% to 20% of this market over the next five years, that will generate recurrent revenues of about 14 M€, with about 40 FTE. Novamet team has a long experience in materials processing and numerical simulations. It has direct access to Academic Institutions and a large network in aluminium industries. A first purchase order from Constellium (first customer) in addition to several grants allowed the development of first industrial scale prototype. This prototype has been installed on a 30-tonnes furnace since February 2019. Two sales contracts are under negotiation.