Transmission tower

Computer Vision model to classify assets on Transmission towers

0.00 credit(s)

Engage a prebuilt deep learning model to analyze defective assets on a transmission tower. The model trained on IBM Maximo Visual Inspection can identify 25 different assets on a transmission tower. 

Enterprises are transforming businesses by leveraging AI methods to quickly analyze health of remote assets for performance and maintenance. Explore IBM Maximo suite of offerings, where complete solution from a video footage from  drones can be monitored for ensuring no loss in productivity.   

Download limit: unlimited
Download availability: no expiration

By adding this digital twin to your cart, you are confirming that you have read the End User License Agreement to this digital twin under the Store Policies tab below.

Add to Wishlist


IBM Maximo Visual Inspection is an award winning training platform designed for subject matter experts to train deep learning models for computer vision.The tool offers a point and click interface to deliver fast, east and accurate models for artificial intelligences. 

Our product is a pre-built deep learning model that can be instantly used to assess health of 25 different assets (read list below)on a transmission tower. The visual insights help supervisors to raise maintenance orders and recommending the type of failure and equipment required to address a fix.

The model represents a deep learning classification model trained by using publicly accessible dataset from Energy Power & Research Institute 

Following assets on a transmission tower are identified by this model:

Conductor Good, Dampers Damaged, Bird Guards, Conductor Damaged, Ground Bonds Broken, Wood Pole Cap Problems, Connectors Corroded, Porcelain Insulators Flashed, No Nest, Misaligned Amor Grips, Misaligned Amor Grips, Nests, Glass Insulators Broken, Polymer Insulators Flashed, Wood Pole Cavities, Misaligned Hardware, Cotter Pin Missing_Loose, Properly Aligned Insulators, Porcelain Insulators Broken, Glass Insulators Good, Polymer Insulators Contaminated, Porcelain Insulators Good, Glass Insulators Contaminated, Misaligned Insulators,  Porcelain Insulators Contaminated and Marker Balls. 

The model provides the following characteristics and the training process can be found here:

  • Number of recognized categories:25
  • Accuracy: 94%
  • Precision: 95%
  • Recall: 95%
  • Network: GoogleNet
  • Framework: Caffe 

Following instructions will get you started:

  1. Download the attached model in this product 
  2. Access an instance of IBM Maximo Visual Inspection by purchasing a copy or downloading our evaluation license or accessing a free 30 day trial
  3. Import the model from this product 
  4. Deploy the model
  5. Validate the model by dragging and dropping images on the User interface
  6. Use one of our code patterns or use the RESTAPIs to write up a cool application  

Please reach out to Srini Chitiveli ( for any help you might need.

Additional information


Digital Twin Resources


Refund Policy

Changes to a non-downloaded order may be available for a period of 14 days from the date of purchase.  Contact the AI Applications Store support at to request support.

Cancellation / Return / Exchange Policy

Changes to a non-downloaded order may be available for a period of 14 days from the date of purchase.  Contact the AI Applications Store support at to request support.

End User License Agreement

General Inquiries

There are no inquiries yet.