African Mining August 2026 | Page 61

MINING COMPANIES WANT 4IR, BUT ARE FAILING AT 3IR

Contributed by Mintek, Edited by Sharon Mdaka
ELECTRA MINING AFRICA •
The mining industry is racing towards automation, artificial intelligence and advanced analytics. Meanwhile, research is showing a less exciting reality.
Background image supplied by Minetek

Research shows a great deal of potential for AI, however, many mineral processing plants still struggle with the fundamentals.

Mintek is actively researching and implementing AI-powered advanced process controllers and is seeing the potential of these technologies in the mining industry. The biggest obstacle to adopting advanced technologies is often not the technology itself, but the condition of the systems that support it. While mining operations have made significant progress in deploying semiautonomous vehicles, collision avoidance systems, and digital monitoring technologies, many processing facilities continue to operate with ageing instrumentation, outdated automation infrastructure, and inconsistent operational data.
“ Mining is adapting quickly – semi-autonomous trucks, collision avoidance systems,” says Dr William Shipman, Research Leader for Fourth Industrial Revolution technologies within Mintek ' s Measurement and Control Division.“ Minerals processing and metallurgy is not moving as quickly, with a lack of research funding into new sensors, old automation systems and slow adoption of modern IT practices. Plants do not invest in maintaining their instrumentation, even when that comparatively small investment can lead to significant improvements in production.”
The result is a growing disconnect between the industry ' s ambitions and its operational reality. Artificial intelligence has become one of the most talked-about technologies in mining. From predictive maintenance and machine vision to advanced process optimisation, suppliers are promoting AI as a solution to improve efficiency, productivity and operational stability.
Mintek warns that many operations may be attempting to skip critical stages in the digital maturity journey.“ We are particularly excited about our Self-learning Artificial Intelligence Optimiser that is getting an extra 2 % throughput on milling circuits with virtually no maintenance,” says David Phillpotts, executive manager of Mintek ' s Measurement and Control Division.“ A major bottleneck to AI adoption persists: many operations have skipped the foundational steps of the information age, making it difficult to transition into advanced digital technologies.”
The challenge is not simply collecting data. It is collecting reliable data. Many processing plants continue to struggle with poor instrumentation coverage, inconsistent measurements and fragmented information systems. Without trusted operational data, advanced analytics and AI tools become significantly less effective. According to Phillpotts, the prerequisites for successful digital transformation remain absent.
“ Many of these emerging technologies originate outside the traditionally conservative process control and automation industries. Consequently, the fundamental prerequisites for stable operations, such as robust measurement, secure information systems and a firm grasp of process dynamics, are frequently lacking. In some cases, the core physics or chemistry of the process itself is neither fully understood nor properly considered.”
One of the more surprising findings highlighted by Mintek is the industry ' s reluctance to invest in instrumentation maintenance. Shipman argues that relatively modest investments in sensors and measurement systems can deliver significant operational benefits, yet these systems are often overlooked during budgeting cycles.
“ Plants do not invest in maintaining their instrumentation, even when that comparatively small investment can lead to significant improvements in production.” The consequences extend beyond production losses. Poor instrumentation directly affects process control, optimisation efforts and the quality of operational decision-making.
For advanced control systems to function effectively, operators must first be able to trust the measurements entering those systems. Without that foundation, digital technologies risk becoming expensive layers of complexity rather than tools that improve performance. •
View the full article online
www. africanmining. co. za African Mining • August 2026 • 59
African Mining Publication African Mining