XXXI International Mineral Processing Congress 2024 Proceedings/Washington, DC/Sep 29–Oct 3 1075
system’s response is highly variable and unstable, resulting
in decreased gains in performance (Traditional). A ‘stabi-
lized’ system is achieved through instrumentation upgrades
or control tuning. However, the system is nowhere near
fully optimized to its full potential (maximum con-
straint). To achieve this, an advanced process control sys-
tem is required, leveraging off process models to push the
operation close to the maximum constraint, which is the
maximum physical operational limit (mechanical\ and
electrical). Enabling a fully integrated AI system requires
the base functional systems (instrumentation) to be fully
operational and optimized. Dynamic constraint adapts to
the changes in the ore type, operational bottlenecks, ESG
(Environment, Social and Governance) KPIs and commod-
ity price changes. Collating, consolidating and analyzing
the breadth of data requires the power of AI systems.
By comparing the maturity levels between automation
and AI adoption, AI systems are suitable at Levels 3 and 4,
which are the automation maturity levels. ERP (Enterprise
Resource Planning) systems use large data sets to plan and
forecast production outcomes (Figure 6). Therefore, at level
4 automation maturity, AI systems become a valuable pre-
diction tool. For example, the AI adoption maturity level
1 starts at the ERP stage of automation—once a mature
automation system is operational.
DEFINING THE VALUE
The value of an integrated mining operation links the
resource definition, mining and process plant constraints
with costs associated with producing the valuable metal
(Figure 7)The objective of the optimized operation is to
increase the efficiency of metal production or minimize the
Figure 6. Automation maturity levels vs AI adoption maturity levels (image adapted from GMG, 2022)
Figure 7. Defining value in an integrated mining operation (Chandramohan, 2023)
system’s response is highly variable and unstable, resulting
in decreased gains in performance (Traditional). A ‘stabi-
lized’ system is achieved through instrumentation upgrades
or control tuning. However, the system is nowhere near
fully optimized to its full potential (maximum con-
straint). To achieve this, an advanced process control sys-
tem is required, leveraging off process models to push the
operation close to the maximum constraint, which is the
maximum physical operational limit (mechanical\ and
electrical). Enabling a fully integrated AI system requires
the base functional systems (instrumentation) to be fully
operational and optimized. Dynamic constraint adapts to
the changes in the ore type, operational bottlenecks, ESG
(Environment, Social and Governance) KPIs and commod-
ity price changes. Collating, consolidating and analyzing
the breadth of data requires the power of AI systems.
By comparing the maturity levels between automation
and AI adoption, AI systems are suitable at Levels 3 and 4,
which are the automation maturity levels. ERP (Enterprise
Resource Planning) systems use large data sets to plan and
forecast production outcomes (Figure 6). Therefore, at level
4 automation maturity, AI systems become a valuable pre-
diction tool. For example, the AI adoption maturity level
1 starts at the ERP stage of automation—once a mature
automation system is operational.
DEFINING THE VALUE
The value of an integrated mining operation links the
resource definition, mining and process plant constraints
with costs associated with producing the valuable metal
(Figure 7)The objective of the optimized operation is to
increase the efficiency of metal production or minimize the
Figure 6. Automation maturity levels vs AI adoption maturity levels (image adapted from GMG, 2022)
Figure 7. Defining value in an integrated mining operation (Chandramohan, 2023)