How Phoenix AI is modernising mining operations

Online Editor

Curtis Stacy reports on the rise of artificial intelligence in blasthole drill automation

As mining companies continue to pursue automation and digital transformation to drive operational excellence, drill and blast crews are ready to adopt AI-based solutions that modernise operations, drive down costs and improve drilling performance.  Deloitte reported in 2018 that “AI-based technologies in the mining industry are here to stay, allowing companies to make more accurate decisions faster, improve health and safety, boost efficiency and ensure human errors are almost negligible, all while helping create smaller environmental footprints.”

This article explores the use of Phoenix AI, the first AI-based blasthole drill automation platform, to optimise drilling and support long-term continuous improvement in areas such as operator performance, hole quality and accuracy, rock fragmentation and blast effectiveness.

Inconsistency issues in mining operations

Today’s blasthole drillers are under increased pressure to maximise penetration and total production without damaging the machine. For machine operators, this requires fast, accurate decisions to continually adjust and fine tune parameters – such as speed, pressure and force – throughout every drilling cycle.

Ground conditions are unpredictable, changing constantly from one hole to the next. Operator performance can be as equally inconsistent from one shift to another. There is also a high margin for human error in detecting geologic faults and manually adjusting drilling parameters to complex, downhole environments.

This lack of consistency in operator performance is one of the highest causes of machine stress and damage, resulting in increased maintenance costs and reduced drill life.

Down-the-hole faults can compromise the entire drilling cycle, leading to poor hole placement, depth and mast alignment. These are all critical factors that have a direct impact on rock fragmentation and blast effectiveness, as well as loading and haulage, and other downstream processes.

Preserving the site’s OEM (original equipment manufacturer) relationship, service and repair model has been another major barrier for crews looking to deploy drill automation platforms. Traditional options to automate drilling options involve replacing the original OEM control system or the installation of a new OEM platform/cab. There are several common challenges to this approach. These include risk to OEM control system, service and support (the OEM control system is removed from the machine to install an automation-enabled platform) and training requirements. Traditional OTM (original technology manufacturer) automation platforms also require significant training requirements in that the entire machine is now controlled differently than it was from the OEM. This requires training for every aspect of the machine maintenance program, from planning down to mechanics in the field.

Supply chain disruption is a further challenge. Service and parts support suffers, resulting in relationship damage and downtime.

Another common issue is that limited Boolean logic is complex and incomplete. Abstract conditions can be difficult to determine with Boolean logic alone. More configuration and tuning is needed as ground conditions change. With Boolean logic, the machine waits for an input signal to exceed a set point before taking action – which is often either too late or the system is tuned to react too soon.

A final challenge is that installation time and costs associated with complete control system replacement are high. Typical installation times for aftermarket automation systems exceed three weeks and cost more than US$100,000 for labour alone to install.

The Phoenix AI edge 

Interoperable with all blasthole drills, of any make or model, the Phoenix AI system empowers the drill operator to automate the entire drilling cycle via an intuitive in-cab dashboard display. The operator can switch easily between autonomous and manual operating mode with the push of a button, without compromising the OEM control system.

The system is intuitive, powered by an AI-based decision engine that continuously monitors and responds to geologic faults in real time. Instead of limited Boolean logic, AI-logic monitors all input signals, reliably determines abstract conditions and takes immediate action to mitigate failure.

The technology determines the design hole depth based on the engineered drill plan and uses advanced AI algorithms to automatically generate and adjust force on the drill bit and rotary RPM to match ground conditions. It takes instant action to correct faults in a nearly predictive manner, can test new drilling parameters and determine positive or negative outcomes based upon that testing. 

Phoenix AI finds the mining operations sweet spot

By employing AI-based drill automation technology mining companies can address common drill and blast challenges. They can eliminate operator variability and automatically mitigate downhole events that can lead to hole failure, machine stress and productivity losses.

Bringing together the best of human and artificial intelligence to optimise the drilling process, Phoenix AI decreases the gap between a good driller and a great driller. It’s like having your best operator in the seat on every shift, combining a layer of machine AI with real-time actionable feedback that eliminates the need for drillers to tune operational parameters on the drill.

By taking instant action to respond to and correct downhole faults, the system ensures the highest quality blast hole. By testing, and learning from, new drilling parameters, the system delivers supports long term continual improvement.

The system ensures accurate execution of the mining engineer’s design and finds the sweet spot for every drill hole – the perfect balance of speed, pressure and force – to maximise penetration rates and total production without damaging the machine. By making live ground condition data available to the entire drill and blast team, they can determine the correct use of explosives relative to the ground conditions for improved blasting and fragmentation.

This technology has shown immediate and sustainable improvements in the field. For example, an Arizona copper mine achieved a 0.30c per foot reduction in drilling costs through the introduction of Phoenix AI drill automation. This has equated to an overall savings of US$210,000 in drilling costs, annually. During the two-month trial period, Phoenix AI increased drill production and penetration rates by 30%. This included a significant improvement in the overall feet drilled and the average drilling and retraction time per hole. For the mine site, this equated to more holes drilled, in less time, without compromising the machine. In a second trial, conducted two months later, Phoenix AI had outperformed all manual drill operations on site with the highest penetration rate.

Phoenix AI reduces machine stress

During the trial, Phoenix AI considerably reduced machine stress without impacting productivity, leading to improved life of component such as MastBits, Steel, Rotary Head, Pumps and Shock subs.

The two figures on page 13 show an increase in stress on mechanical components of the drill (the red and orange lines) even while providing lower productivity (blue line). The operator does not adjust the drill parameters (primarily feed force) often while drilling where the auto drill constantly adjusts to match ground conditions.

Key observations are: the operator exceeds the maximum penetration rate and risks getting stuck in the hole; as the penetration risk spikes, Phoenix AI autonomous mode reduces the force on the bit while manual mode makes no adjustments; the operator drills at near maximum parameters, causing undue stress on the machine.

Mining operations performance rates improves

Blasthole drills were 12% more productive in autonomous mode than the most productive driller operating manually. Compared across three drills and all operators, the Phoenix AI delivered the highest performance and penetration rate of all three.

Hole quality is one of the most notable improvements of AI-based drill automation. Correcting downhole issues during the drilling cycle results in high quality holes that play a key role in ensuring optimised fragmentation and blast outcomes.

The future of mining operations

Drill and blast teams seek interoperable, integrated digital technologies that can improve decision making and collaboration at multiple levels. AI is enabling greater insight than ever before and will continue to lead the way in optimising equipment and processes across the entire mining value chain.

AI-optimised technologies help mining operations improve efficiency and productivity, by making faster, more accurate decisions and reducing human error to optimise machine performance.”

Curtis Stacy is with MineWare