Chinalco Launches AI-powered Real-time Online Analysis Application for Bauxite Composition

Recently, Chinalco Group launched its independently developed “Real-time Online AI Analysis Application for Bauxite Composition”, which efficiently realizes real-time online detection of components such as alumina, ferric oxide and silicon dioxide in bauxite. The single detection time has been shortened from 4 hours to 20 seconds, providing strong technical support for scientific and reasonable ore blending and stable and efficient operation of alumina production, Xinhua News Agency reported.

With the increasing tightness of global bauxite supply, the origins of ore used in alumina production have become diverse and the composition fluctuations have intensified, directly affecting product quality and production stability. At present, alumina enterprises generally adopt the traditional “manual sampling + laboratory sample preparation and analysis” method for ore detection, which has a long process and slow feedback and cannot meet the demand for precise ore blending.

Developing online detection technology and equipment with real-time response capability has become an urgent task to promote the intelligent upgrading of the alumina industry and improve production efficiency and stability. China Economic Net noted that intelligent detection technology has become a key driving force for the digital transformation of the non-ferrous metal industry.

To address this challenge, Chinalco Group has systematically studied bauxite spectral data collection methods, innovatively adopted preprocessing technology integrating data and physical mechanisms, and built a high-accuracy deep AI detection model driven by feature fusion and transfer learning. It has also successfully developed bauxite online detection equipment with all core components localized, filling the gap in real-time bauxite detection at home and abroad.

Zhang Hong, chief engineer of Chinalco Group’s Technology Research Institute, said the application embodies the group’s technological strength in intelligent transformation. “As a leader in the non-ferrous metal industry, Chinalco has been deeply integrating AI technology into the entire industrial chain, which is in line with the industry’s development trend of digital and intelligent upgrading,” he stated.

At present, this achievement has been industrially applied in two alumina enterprises. While the detection efficiency has increased by 720 times, the average relative error of various component detections in bauxite is less than 5%. It can adapt to the composition changes of various types of bauxite at home and abroad, and can real-time guide alumina enterprises to carry out precise regulation, significantly improving the continuity and stability of alumina production with prominent economic and social benefits.

Xinhua News Agency added that Chinalco has been promoting the application of its self-developed “Kun’an” AI large model in the non-ferrous metal industry, and this bauxite online analysis application is an important practice of the group’s digital transformation strategy, contributing to the high-quality development of the alumina industry.