Let’s talk
Contact us
Particle Size Distribution Analysis
Real-time particle size analysis for bulk materials. Powered by AI and Machine Vision for precise detection and classification, with accuracy to 0.0125 mm.
The CVST ML Sense Platform
An AI-driven platform for automated quality control on conveyor production lines. Pre-configured solution within the ML Sense platform, designed to:
Detect
Detect individual particles in a live video stream.
Classify
Classify them by size in real-time.
Identify
Identify oversize material and foreign contaminants.
The Limitations of Traditional Methods
Traditional particle analysis is plagued by:
  • Inaccuracy
    Subjective methods can have error rates up to 80%.
  • Inefficiency
    Slow, manual sampling and lab analysis.
  • Delay
    Results take hours or days, preventing real-time process adjustment.
  • Safety Risks
    Increased manual handling in hazardous areas.
The Critical Role of Particle Size Analysis
Particle size is a direct indicator of efficiency throughout your mining and processing operations.
Oversized Rocks
Reduce bucket fill factors and can cause catastrophic damage to crushers, leading to costly downtime.
Excessive Fines
Often signal overuse of explosives during blasting, representing a significant and unnecessary expense.
The cost of inaccurate analysis isn't just data—it's production halts and lost revenue.
The ML Sense system delivers objective, real-time particle size analysis with up to 95% accuracy, even in the most demanding industrial environments.
Core Capabilities of ML Sense
  • Live Particle Detection
    Our system continuously identifies and monitors individual particles within a live video stream of your bulk material flow.
  • Automated Size Calculation
    It instantly calculates the size of each particle using the area method for precise, objective measurements.
  • Intelligent Particle Classification
    Particles are automatically sorted into your pre-defined categories based on their calculated size.
  • Real-Time Fraction Analysis
    The module provides immediate reporting on the percentage composition of fine, coarse, and blended fractions.
  • Contaminant & Oversize Identification
    The system automatically flags oversize material, tramp metal, and other foreign objects for immediate action.
  • Comprehensive Data Logging
    All measurement data and events are systematically recorded for in-depth reporting, analytics, and process optimization.
See It In Action

  • Fraction Analysis

    Live reporting of fine, coarse, and mixed fractions on the conveyor belt.

  • Oversize Detection

    Automatic alerts when particles exceed predefined size thresholds.

  • Data & Reporting

    A clear, intuitive event log with full data export capabilities for any period.

Optimize Your Mining Workflow
For Processing & Crushing
Prevent crusher jams and downtime by automatically identifying oversized rocks before they reach the crusher. Optimize crusher settings in real-time based on the actual feed composition.
For Drill & Blast Operations
Fine-tune blast patterns and explosive use based on post-blast fragmentation analysis. Use the mobile version on a tablet or smartphone directly at the blast site for immediate feedback.
Technology Stack
Object detection
Yolo, SSD-MobileNet, SSD-Inception
Image segmentation
PixelLib, Mask R-CNN, U-net
Complex CV Algorithms
OpenCV, Scipy, Numpy, Scikit-learn
Self-contained Devices
Nvidia Jetson nano, Nvidia Jetson AGX Xavier, Google Coral
Intelegent cameras
Intel realsense, Hikrobot Smart Cameras
Latest algorithms
Reinforcement Learning, StableDiffusion
Get Started ML/AI Project

What our clients say about us

Mikhail P.
Quality Manager in Ariston Thermo
Ariston Thermo would like to thank CVST for their outstanding work in integrating an automated defect detection system for our assembly line.
The team at CVST demonstrated deep expertise in machine learning and seamless collaboration at every stage—from initial design to final implementation. The system is now operating with high accuracy and stability, effectively identifying defects and providing us with valuable data for process optimization.
We were particularly impressed with their professional approach and prompt resolution of any issues. The results have fully met our expectations, and we highly recommend CVST for their expertise and reliable partnership.
Mikhail P.
Quality Manager in Ariston Thermo
Ariston Thermo would like to thank CVST for their outstanding work in integrating an automated defect detection system for our assembly line.
The team at CVST demonstrated deep expertise in machine learning and seamless collaboration at every stage—from initial design to final implementation. The system is now operating with high accuracy and stability, effectively identifying defects and providing us with valuable data for process optimization.
We were particularly impressed with their professional approach and prompt resolution of any issues. The results have fully met our expectations, and we highly recommend CVST for their expertise and reliable partnership.
Olga T.
Head of Human Resources
Based on the testimonial from Nissan Manufacturing, CVST's NissanBot Telegram solution has been a resounding success. Initially an "electronic doctor" for staff, it now serves nearly 90% of employees as a vital communication hub, especially without corporate email.
The bot streamlined HR processes, automates surveys, and provides instant access to work information. Its broadcast feature became a unified channel for all corporate messages, growing from 200 to 1800 users in a month. The platform ensures data confidentiality.
90% of users find the bot convenient, and 60% actively use it for news. The admin panel is praised for its simplicity, backed by CVST's responsive and helpful support team. A highly effective and essential internal tool.

Contact us

Please share your contact details as well as project info
Attachments
By clicking the button you agree to our Privacy Policy