Our Clients
Espy Computer Vision System
Ore Properties
The Espy Computer Vision System (ECVS) is capable of detecting ore properties ranging from Particle Size Distribution, Grade, Oversize and Crushing Index. The ECVS comprises industrial grade Machine Vision Cameras or standard CCTV cameras enhanced with 3D scanner, LIDAR technology or ultrasonic level sensors to provide the richest information for the state of the art convolutional neural networks specifically designed for the application.
Learn More
Blockages Control
The number one downtime in mining processing is blockages. Espy Systems offers a comprehensive solution to predict and prevent or to detect and control blockages at unprecedented efficiency. Our solutions use the ECVS vision system to detect bridges and blockages in crushers, hoppers or bins, on top of screens. The ECVS interacts with the local process control system to control feeders, rock breakers or conveyors in order to prevent or mitigate blockages.
Learn More
Contamination
The Espy Computer Vision System (ECVS) can detect most contaminants visible in loader buckets, truck trays, hoppers or bins or on running equipment, conveyors, screen, flotation cells etc. The ECVS utilises the fused data of Machinve Vision Cameras and Depth Sensors to recognise and locate any contamination ranging from oversize, timber, roof bolts, liquids etc. The ECVS interfaces with the process control system to coordinate the screening or reject processes.
Learn More
Espy - Computer Vision meets Process Optimisation
-
End-to-End Process Control Solution
Espy is a cutting edge software solution built to combine the power of Advanced Process Control, Computer Vision and AI.
-
Plant Data Analytics and Optimisation
Machine Learning, Big Data analytics and Computer Vision combined powers the most advanced algorithms to maximise process performance.
-
-
Computer Vision
Visual data is the richest, readily available and least utilised data on most mining sites providing great potential for the Espy Computer Vision System to realise value. Espy uses specialised software to analyse live footage from any cameras to detect objects, material properties or anomalies.
-
Deep Learning
A unique feature of Espy is its learning capability, a form of Artificial Intelligence that mimics the pattern recognition capabilities of humans and surpasses conventional non-linear modeling techniques. Espy is built on state-of-the-art neural network architecures
The Power of Edge Computing
-
Industrial Edge Computing
The Espy AI Edge Controller is a fanless, multiple CPU and GPU equipped, industrial edge computer specially designed for heavy AI loads. It has multiple ethernet and rich IO interfaces. The Espy AI Controller is based on the same hardware platform than the one used for autonomous driving. We offer regular software and hardware upgrades as technology advances, which has the added benefit of featur expansion, increase in accuracy and processing speed.
-
Locally hosted, remotely supported
The ECVS is a locally hosted, stand-alone system that offers the fastest possible response and realtime processing while also maintains the highest level of security and IT segregation. All the pipelines required for commissioning and maintenance are routed through the local IT networks in order to provide security and reduce deployment and maintenance costs.
Computer Vision System built on top of cutting edge technologies
Computer Vision
-
CCTV Cameras
The ECVS can interface with site standard CCTV cameras that are perfectly suitable for a number of applications. While they might not be ideal for most of the computer vision applications there are a number of advantages utilising well-maintained, good quality CCTV cameras. When restrictions on resolution and image response lag are not compromising factors the ECVS can be integrated with one or more CCTV cameras. The most common applications are crusher or ROM bin bridge detection, ore grade recognition, flotation cell or screen efficiency optimisation etc.
-
Machine Vision (MV) Cameras
MV cameras do not use compression therefore they require higher bandwidth and dedicated network architecture (at least 1Gb/s network switches). The camera and lens design are unique to the given application and target view. There are a number of environmental and application specific factors considered and translated to exposure time, resolution, frame rate, dynamic range etc. The ECVS for moving object detection requires a high sensitivity, high speed sensor with global shutter and low exposure time. The result is a specially designed camera the offers low latency, high resolution, high speed, crisp images.
Applications
-
Particle Size Distribution
Real-time detection of ore particle sizes, detection of oversize material, lump in fines and contamination detection, crushing, milling and screening efficiency measurement and optimisation.
-
Blockage and Bridge Control
The number one downtime in mining processing is blockages. Espy Systems offers a comprehensive solution to predict and prevent or to detect and control blockages at unprecedented efficiency. Our solutions use the ECVS vision system to detect bridges and blockages in crushers, hoppers or bins, on top of screens. The ECVS interacts with the local process control system to control feeders, rock breakers or conveyors in order to prevent or mitigate blockages.
-
Enrichment Process Optimisation
Multivariable chemical dosing and mixing processes benefit immensly from high spectrum, online visual analytics done by the Espy Computer Vision System. State of the art global optimisation and multivarible control techniques model these complex enrichment processes and improve recovery rates far exceeding the performance of expert systems and human operators.
-
Belt Damage Inspection
Predictive maintenance, online non-contact monitoring of conveyor belt drift, side loading, wear and damage. The Espy Computer Vision System (ECVS) uses visual data, which is has far greater information density than conventional acoustic and radar based belt monitoring systems. The visual data is processed by state-of-the-art deep neural networks that can detect anomalies with extremely high accuracy even beyond human capabilities.
-
Contamination Detection
Detection of conveying fines product anomalies such as liquefaction, lump in fines, foreign material in ore streams like timber, liners, mill balls etc.
-
Ore Profile Detection
Online measurement of the ore profile, edge distance and volumetric belt loading can help to maximise the throughput of conveyor belts, detect surges, feeding and tracking issues, eliminate bunching and side loading.
ESPY - Advanced Automation through Intelligent Instrumentation
Espy Data Sciences - Expert Data Engineering fused with Rapid Implementation