AI V-CORE SANBOS V-CORE Platform

Enabling Machines to See
the World Like Humans Do.

Next-generation industrial robotics, powered by vision at the core. We use AI perception to reshape the control loop and accelerate the intelligent evolution of multi-form robots. Sanbo AI’s SANBOS V-CORE Platform is vision-centric—an integrated edge-to-cloud AI platform for PLC guidance, robotic bin picking, and few-shot defect inspection: the vision brain for robots.

AI for PLC Intelligent Guidance: V-CORE AI redefines industrial control logic—no point-by-point programming, reducing PLC complexity and commissioning time while raising line efficiency.

V-CORE Robotic Bin Picking: V-CORE AI vision and motion planning address misalignment, recognition issues, and lengthy commissioning in material handling—enabling flexible, stable, highly adaptable automated feeding and sorting.

V-CORE Cloud-Edge Few-Shot Defect Inspection: Few-shot learning overcomes large sample and labeling costs—enabling fast model training and high-precision, easy-to-deploy industrial quality inspection.

SANBOS V-CORE Platform VISION CORE
Edge Box
Real-time inference · PLC / robot control
Cloud Platform
Model hub · inference · decision · planning
Robots
Guidance, bin picking, environment perception
V-CORE Vision-Guided Robotic Bin Picking
V-CORE Cloud-Edge Few-Shot Defect Inspection
PRODUCTS

Products & Solutions

Built on the SANBOS V-CORE AI vision engine, we use AI-powered perception to reshape the control loop. We address the limitations of traditional PLC control—rigid logic, tedious programming, long development cycles, and difficult on-site commissioning—with flexible, scalable AI vision that enables autonomous robot guidance without custom code. Combined with V-CORE few-shot learning, we deliver rapid model training and industrial inspection solutions that are accurate, easy to operate, and quick to deploy for integrated smart manufacturing.

AI for PLC Guidance

Replace rigid PLC logic with AI vision so PLCs perceive part position and state—reducing commissioning cost and enabling flexible production.

  • Compatible with mainstream PLCs and fieldbus protocols for drop-in integration with existing lines
  • Vision-driven guidance instead of hard-coded logic—less rewiring and trial-and-error
  • Unified positioning and defect inspection to reduce duplicate systems

V-CORE Robotic Bin Picking Guidance

3D perception, pose estimation, and path planning for reliable picking from unordered bins.

  • Support for multiple camera and robot combinations with fast deployment across workstations
  • Robust to occlusion and reflection in real factory environments
  • Integration with MES and PLCs for closed-loop automation from feeding to unloading

V-CORE Cloud-Edge Few-Shot Defect Inspection Platform

Edge devices run millisecond-level inference; the cloud manages models and continuous improvement so every line improves over time.

  • Centralized cloud management of models and config with one-click rollout to edge devices
  • Online updates and phased rollout to minimize line downtime
  • Data feedback and labeling loop for ongoing algorithm evolution
TECHNOLOGY

Architecture & Cloud-edge Collaboration

Based on the V-CORE edge node and cloud platform, we deliver an “edge inference, cloud decision and planning” architecture. From industrial cameras to the cloud, every layer is designed for reliability and operability in real factory environments.

Edge & Camera

  • Support for multiple industrial cameras, 3D cameras, and lighting
  • Edge compute boxes with GPU / FPGA and multi-stream concurrency
  • Secure connection to PLC / robot control cabinets

Cloud Platform

  • Model versioning and A/B experimentation
  • Unified monitoring across production lines and stations
  • Data feedback, labeling, and retraining interfaces

Cloud-edge Topology

  • Cameras / sensors → edge nodes → PLC / robots
  • Cloud as model and config hub with per-station rollout
  • Edge runs independently so lines keep running if the cloud is briefly unavailable
  • Data sent back to the cloud as needed for offline analysis and model iteration

Operations

  • Cloud-edge heartbeat monitoring and alerting
  • Remote configuration and log collection
  • Metrics: yield, cycle time, downtime causes, and more

Replace this area with your network topology diagram or product screenshots when deploying.

SANBOS V-CORE Platform

From Machine Vision to Robot Visual Brain

With SANBOS V-CORE as our technology foundation, we focus on VISION CORE capabilities built on deep learning and reinforcement learning—delivering vision-guided services for robots and making “see clearly” a standard capability for smart manufacturing.

Product 1
AI for PLC Intelligent Guidance System

Replace traditional rigid logic with AI vision so the PLC perceives workpiece position and state—reducing commissioning cost and improving flexible production efficiency.

Product 2
V-CORE Robot Random Bin Picking Guidance

For randomly stacked parts, we provide 2D-instead-of-3D visual recognition, pose estimation, and trajectory planning for stable, efficient robot picking.

Product 3
V-CORE AI Few-Shot Defect Inspection Platform

Industrial-grade AI vision for high-accuracy defect detection and fast model deployment with minimal labeled data.

TEAM

Team & Background

Co-founded by PhDs and postdocs from Fudan University and rooted in Fudan’s Institute of Intelligent Robots and Advanced Manufacturing, we bring together executives from multinational and listed companies to build a world-class AI team and deliver a deployable vision brain for robots.

Founding Team

Our development team includes PhDs, postdocs, and professors with deep expertise in machine vision, robot control, deep learning, and reinforcement learning—backed by strong academic and engineering execution.

Industry Experience

Senior executives from multinational and listed companies join as co-founders, combining large-scale R&D and production experience with academic innovation to deliver practical, deployable products.

Mission

With vision as the core engine, we use AI perception to reshape the control loop and drive the continuous evolution of multi-form robots!

SCENARIOS

Typical Application Scenarios

Representative applications built on AI for PLC guidance, V-CORE bin picking, and V-CORE few-shot defect inspection—helping production lines become more flexible and intelligent.

Flexible Assembly & Line Guidance

With AI for PLC intelligent guidance, the same line can switch product variants quickly with minimal downtime and commissioning—improving order flexibility and enabling low-code or no-code vision guidance.

Parts Bin Picking

Using V-CORE robotic bin picking guidance, we deliver stable 3D localization and picking from parts bins in automotive, 3C, and other industries—replacing manual sorting and improving feeding efficiency and safety.

Multi-station Few-Shot Defect Inspection

Using the V-CORE cloud-edge few-shot defect inspection platform, cameras and edge nodes are deployed across stations; the cloud manages models and inference, decision, and planning—continuously improving yield and cycle time with data-driven analytics.

FAQ

Frequently Asked Questions

  • How do you integrate with existing PLCs and production lines?
    We integrate with mainstream PLCs and fieldbus protocols. The V-CORE edge node typically connects to PLC cabinets via digital I/O or industrial communication interfaces for guidance and status signals, enabling smooth integration with existing lines.
  • How much data does V-CORE few-shot defect inspection need?
    It depends on defect type; we provide recommendations after an on-site assessment. The V-CORE platform supports few-shot learning—in many scenarios you can train and deploy with limited labeled samples, then improve continuously through data feedback.
  • Is deployment cloud-based or on-premises?
    Defect inspection and guidance inference run on the V-CORE edge for real-time, stable performance. The cloud handles the model hub, inference and decision planning, configuration rollout, and multi-station monitoring.
  • How much data is needed, and how long until go-live?
    We provide data collection and labeling guidance for each scenario. The V-CORE cloud-edge few-shot defect inspection platform lets you start with limited data and improve continuously after deployment through data feedback and labeling loops.
CONTACT

Contact Us

We welcome production line managers, automation integrators, and partners to explore SANBOS V-CORE for line upgrades and collaboration—whether for AI for PLC guidance, V-CORE vision-guided bin picking, or the V-CORE cloud-edge few-shot defect inspection platform. We look forward to helping you bring vision to robotics.

Company / Organization
Name
Email
Phone
Message
A team member will contact you after submission. Thank you for your interest and support!
Direct Contact
Email: baorui@sanboai.com
Phone: +86 139 1143 2980
WeChat / Work WeChat QR WeChat / Work WeChat
Official Account QR Official Account

Beijing HQ1F, Block A/B, Dongsheng Building, 8 Zhongguancun East Road, Haidian District, Beijing

Shanghai589 Tongbei Road, Pingliang Road Subdistrict, Yangpu District, Shanghai

Careers (Beijing, Shanghai)

AI Market Partner

Market development, channel and ecosystem partnerships for industrial vision and robotics products under the vision-centric SANBOS V-CORE Platform; support regional and industry deployment.

Requirements:
  • Bachelor’s degree or above; 5+ years in industrial automation, smart manufacturing, or AI product market development;
  • Channel and ecosystem experience; experience with To B key accounts or system integrators;
  • Understanding of industrial vision, robotics, PLC; able to drive regional/industry deployment independently.

Marketing Engineer

Promotion and marketing of SANBOS V-CORE Platform products including AI for PLC intelligent guidance system, V-CORE robots, and V-CORE cloud-edge few-shot defect inspection platform; customer maintenance, technical exchange, and pre-sales support.

Requirements:
  • Associate degree or above in engineering or marketing; 3+ years in To B technical marketing or pre-sales support;
  • Strong communication skills; able to develop and maintain customer relationships independently.
  • Familiarity with industrial vision, PLC, or robotics preferred; smart manufacturing / automation industry experience preferred.

Algorithm Engineer

R&D and engineering for machine vision and robot control; support SANBOS V-CORE Platform product deployment.

Requirements:
  • Master’s or above in CS, automation, math, physics, or related; solid programming and math foundation;
  • Experience with deep learning / reinforcement learning; machine vision, object detection, or robot control projects;
  • Proficient in Python/PyTorch/TensorFlow; C++ engineering and edge deployment experience preferred.

Scan to add WeChat

WeChat QR code
ContactUs +86 139 1143 2980
TOP