XXXI International Mineral Processing Congress 2024 Proceedings/Washington, DC/Sep 29–Oct 3 1077
for the mining industry does not exist. Therefore, a bespoke
AI-driven system is needed to solve the complex metals
value chain -tracking the metal from the mine to the pro-
cess plant to the market and finally to the manufacturers
and recycling industries. Such a system potentially needs to
leverage the concepts of Blockchain, where each unit metal/
commodity can be tracked from the pit to the final manu-
factured product through a digital fingerprint (Ellis, 2022).
Regarding relative value-add to underperforming min-
ing operations, the most significant gains in increased
productivity are typically just by doing the basics. These
include improving the process control, operating under
optimized operating conditions to achieve stability, and
therefore allowing the mining operations to push against
the maximum constraint of the system. Once a stable oper-
ation is achieved, technology drivers such as AI systems,
real-time monitoring and optimization become viable solu-
tions, Figure 10.
For example, implementing an AI-driven optimization
system for a SAG mill operation requires an advanced con-
troller, such as APC or MPC, as a minimum, Table 2. The
most significant gains (in throughput and comminution
efficiency) are from implementing the APC or the MPC,
where the SAG mill optimally operates close to the maxi-
mum constraint. Finally, an AI-driven solution builds on
the APC/MPC system operating to the dynamic constraints
of efficiency parameters such as energy and consumption,
operating costs, and ESG metrics. The AI-driven controller
Table 1. Defining the value objective for each segment of the mining industry
Operating Mine Developing Mine Environment &Community
Value Objective Maximise throughput,
Maximise recovery
Identify higher grades in the
resource
Identify deleterious elements
Minimise energy and water
consumption in constrained areas
Minimise mine-footprint on the
environment
Maximize community engagement
(jobs, governance)
Metrics of performance Throughput and metal recovery $/t and Net-present-value Reduced GHGe
Permitting
Profit share
Measurable data time
frames
Instantaneous/real-time data
from sensors
Inferred data from drill samples Survey or geological data regional laws
or governance
Data accuracy/precision Real-time, instantaneous data
Highly precise
Accuracy based on statistical
models inferring the drill hole
assay data
Highly dependent on the mine life and
other external factors (environment,
social impact, governance)
To whom is the value
definition aimed?
The mine operator Shareholders of the mining
company
Environment, Community and
regional laws
Figure 9. An AI-driven mine-to-market vision
for the mining industry does not exist. Therefore, a bespoke
AI-driven system is needed to solve the complex metals
value chain -tracking the metal from the mine to the pro-
cess plant to the market and finally to the manufacturers
and recycling industries. Such a system potentially needs to
leverage the concepts of Blockchain, where each unit metal/
commodity can be tracked from the pit to the final manu-
factured product through a digital fingerprint (Ellis, 2022).
Regarding relative value-add to underperforming min-
ing operations, the most significant gains in increased
productivity are typically just by doing the basics. These
include improving the process control, operating under
optimized operating conditions to achieve stability, and
therefore allowing the mining operations to push against
the maximum constraint of the system. Once a stable oper-
ation is achieved, technology drivers such as AI systems,
real-time monitoring and optimization become viable solu-
tions, Figure 10.
For example, implementing an AI-driven optimization
system for a SAG mill operation requires an advanced con-
troller, such as APC or MPC, as a minimum, Table 2. The
most significant gains (in throughput and comminution
efficiency) are from implementing the APC or the MPC,
where the SAG mill optimally operates close to the maxi-
mum constraint. Finally, an AI-driven solution builds on
the APC/MPC system operating to the dynamic constraints
of efficiency parameters such as energy and consumption,
operating costs, and ESG metrics. The AI-driven controller
Table 1. Defining the value objective for each segment of the mining industry
Operating Mine Developing Mine Environment &Community
Value Objective Maximise throughput,
Maximise recovery
Identify higher grades in the
resource
Identify deleterious elements
Minimise energy and water
consumption in constrained areas
Minimise mine-footprint on the
environment
Maximize community engagement
(jobs, governance)
Metrics of performance Throughput and metal recovery $/t and Net-present-value Reduced GHGe
Permitting
Profit share
Measurable data time
frames
Instantaneous/real-time data
from sensors
Inferred data from drill samples Survey or geological data regional laws
or governance
Data accuracy/precision Real-time, instantaneous data
Highly precise
Accuracy based on statistical
models inferring the drill hole
assay data
Highly dependent on the mine life and
other external factors (environment,
social impact, governance)
To whom is the value
definition aimed?
The mine operator Shareholders of the mining
company
Environment, Community and
regional laws
Figure 9. An AI-driven mine-to-market vision