THE BIGGEST CHALLENGES WITH AUTOMATED INSPECTION ARE
DELIVERING ACCURATE, RELIABLE AND COST EFFECTIVE SOLUTIONS

Roadblocks that prevent success

Complex Parts & Surfaces

Final visual inspection of complex parts is incredibly challenging to automate because of part geometries, variations in acceptable surface finishes, and region specific inspection tolerances.

Scarcity of Data

Data is a solution requirement. Complex parts often have a multitude of potential defect classes, with some being very rare. A scarcity of data typically leads to increased project delays and poor performing systems.

System Performance

Final system performance will ultimately impact the ROI or success achieved by the customer. Customers need solutions that work consistently with minimal support or interventions required.

A unique technology developed by Musashi AI

An architecture of two independent algorithms  operating and communicating information to manage complex quality inspection. High performing AI with low initial data requirements. A winning combination of advanced technology for customer applications anywhere.

Our Active i algorithms:

Anomaly Detection

Active i can be deployed with data taken only from defect free part samples. Defects or anomalies will be detected immediately with the anomaly detection algorithm, and these detections enable training and optimization of the instance segmentation algorithm.

Training Data Required
20-30 Defect Free Part Samples
Defect Detection
Yes
Defect Classification
No
Gauging or Measurements
No

Trained with Defect Free Parts

Heatmap of Anomalies

Instance Segmentation

Training of this algorithm begins with the available defect data at initial deployment. With iterations, this algorithm is optimized for improved detections and defect classification. Assembly verification and gauging checks are also managed by this algorithm.

Training Data Required
50-100 Defective Samples per Defect Class
Defect Detection
Yes
Defect Classification
Yes
Gauging or Measurements
Yes

Trained with Defect Free Parts

Detects and Classifies Defects

Technology value:

Active i accelerates the delivery and improves the overall performance of a working automated vision solution. It was designed specifically to solve real problems and ultimately deliver customer success!

Deployment Challenge

Other Solution Providers

Active i from

Complex Parts & Surfaces

A solution is delivered that is not capable of complete part inspection

  • Incomplete or manual inspection required
  • ROI or Cost Savings are not achieved

A solution that is capable of completepart inspection

  • Fully automated inspection
  • ROI or Cost Savings are not achieved

Scarcity of Defect Data

Collect additional customer data

  • Delays deployment and adds to project costs


Data unavailability causes poor system performance

  • Low accuracy and high number of false calls

Low initial data requirements

  • Train and deploy with available data

Iterative model optimization

  • Two algorithms managing inspection delivers high accuracy and low number of false calls

System Performance

Experiment with other approaches

  • Delays deployment and adds to project costs.


Poor system performance leads to costly interventions

  • Customer success is low

Proven performance with the active i architecture

  • Confident deployment of a solution

Accurate and reliable system operation

  • Customer success is high

Complex Parts & Surfaces

Other Solution Providers

A solution is delivered that is not capable of complete part inspection

  • Incomplete or manual inspection required
  • ROI or Cost Savings are not achieved

Active i  from Musashi AI.  A solution that is capable of completepart inspection

  • Fully automated inspection
  • ROI or Cost Savings are not achieved

Scarcity of Defect Data

Other Solution Providers

Collect additional customer data

  • Delays deployment and adds to project costs

Data unavailability causes poor system performance

  • Low accuracy and high number of false calls

Active i  from Musashi AI.

Low initial data requirements

  • Train and deploy with available data

Iterative model optimization

  • Two algorithms managing inspection delivers high accuracy and low number of false calls 

System Performance

Other Solution Providers

Experiment with other approaches

  • Delays deployment and adds to project costs

Poor system performance leads to costly interventions

  • Customer success is low

Active i  from Musashi AI.

Proven performance with the active i architecture

  • Confident deployment of a solution

Accurate and reliable system operation

  • Customer success is high

Talk to an expert or request a trial of Active i for your vision application.