CANADIAN TECHNOLOGY
iteration of this machine , the S20 , was used with success on four shaft projects in both soft rock – with an integrated tool package – and hard rock .
The S20 + offers similar capacity / productivity to a North American pneumatic telescopic boom mucker but with a much easier learning curve for operators , according to Kevin Melong , Vice President – Shafts and Technical Services , Redpath Canada .
Unlike the RD S100 , which can move independently of the galloway within the shaft via a telescopic boom , the S20 + is connected to the galloway , requiring the lowering of the galloway to move the mucker into position for mucking .
Melong added : “ The S20 + does present some potential for concurrent mucking when in and around the shaft station construction / excavation , but , in general , does not act independent of the stage as in the S100 design .”
Fitted with a bucket as standard , the S20 + offers a maximum digging depth of 7.96 m and a bucket volume of 560 litres . It also offers good flexibility , with the bucket capable of six rotations per minute .
Alongside the S20 + and the S100 , Redpath has been mechanising the shaft sinking process in mining using large-diameter raiseboring equipment and operation of Herrenknecht ’ s Shaft Boring Roadheader ( in soft-to-medium rock ). The company is also working with Herrenknecht on the development of the Shaft Boring Cutterhead , a machine designed for medium-to-hard rock applications up to 250 MPa UCS .
Cascadia enters ML-backed , real-time territory
Vancouver-based Cascadia Scientific , known for its application of machine-learning models for optimising mine haulage , is now entering the arena of short interval control with its new “ ML Coach ” suite of tools .
ML Coach is powered by boosted “ tree models ” trained on thousands of extensively characterised site-specific haul cycles , the company explains . These models enable mine staff to dissect every element of their haulage to understand and target those which represent the greatest opportunities for improvement .
Among these elements , ML Coach directly determines the impacts of individual trucks and specific operator performance on fuel efficiency and cycle time . It is also designed to deliver these insights with dispatch-style user interfaces fed by real-time data streams in a matter of minutes .
Kevin Dagenais , CEO of Cascadia , said this real-time data insight represents a significant improvement on ML Coach ’ s predecessor solution .
“ Before ML coach , we were doing very similar
ML Coach ’ s “ tree models ” – trained on thousands of extensively characterised sitespecific haul cycles – enable mine staff to dissect every element of their haulage to understand and target those which represent the greatest opportunities for improvement
work or analysis , but all of this was retrospective – findings were made after the fact , off-line , days or weeks after the activity was competed ,” he told IM .
The company ’ s largest customer at the time – a large coal miner – was successfully using these tools to target operator re-training and also to detect poorly performing haul trucks , with information then sent to maintenance , often with very material findings impacting the performance , according to Dagenais .
“ Our customer challenged us in mid-2022 , explaining that to take the next major step ( expansion from roughly a quarter of the fleet to 100 %), we would need to build real-time tools that could be used for direct intervention in the moment ,” he said . “ This was the impetus for starting our journey towards a real-time ( short interval control ) offering that could bring a real time user model to our existing retrospective tools .”
ML Operator Coach – the latest iteration – maintains a model based on past mine activity that can predict the fuel and time required to complete any haul cycle , according to the company . The model is trained on properties including distance , tonnage , vertical travel , duration and even dump and load locations .
Immediately upon the completion of a haul cycle , ML coach compares the measured results with the model prediction , with the difference between the two values representing a gain or penalty . Over the course of a shift , operations staff monitor the leaderboards to offer support and guidance to operators , when they consistently underperform based on the model .
“ ML coach also identifies operators worthy of praise and recognition for those outperforming model predictions ,” the company says . “ Overlaid
on this are any detections of discrete behavioural events , such as travel at excessive speed and improper application of the service brake . The short interval approach means that , for the first time , tactical human behaviours negatively impacting haulage efficiency can be addressed in real time .”
The second pillar of the ML Coach suite of tools will focus on terrain management . Slated for release in the first half of 2023 , ML Terrain Coach will store models capturing the expected performance of a network of haul road segments , including statistics on expected speed , fuel intensity , vibration and grade .
As instrumented vehicles transit these segments , the measured performance is compared with model predictions . The tool will alert mine staff when road segments drift below threshold key performance indicators . Road condition events , including racking and impact detections , will be displayed as a map overlay .
ML Coach requires a data pipeline with much lower latency , according to the company . This challenge was addressed with upgrades to Cascadia Scientific ’ s on-board hardware and firmware offerings . These provide global connectivity on next-generation machine-tomachine mobility networks including Narrowband IoT and options for accessing existing private LTE or in-pit Wi-Fi .
Cascadia Scientific is also developing short interval control tools and strategies for dozer optimisation in collaboration with a major Canadian oil sands producer .
Automating aerial monitoring
Canada-based Exyn Technologies has been making major inroads in the mining space through the rollout of its ExynAero™ platform for drone-based missions .
It recently partnered with Easy Aerial and its Osprey platform to launch its new flagship airframe , the ExynAero™ EA6 , integrated with ExynAI™ and optimised for Exyn ’ s diverse use cases .
58 International Mining | APRIL 2023