1. EN
  2. FR
  3. ES
  Back
Aug 23 2021

Better operational decision-making through machine learning

Posted by TC Energy

TC Control Room 1

TC Energy is proud to be an energy industry innovation leader that seeks to implement best practices and new technologies to continually improve the safety, efficacy and cost-efficiency of its operations.

Transitioning skillsets to a younger workforce
With an aging trades workforce and a need to transition skillsets to younger generations, there is an opportunity to create ways to provide everyone, whether they have been in the position for 30 years or one month, with a more consistent knowledge base for their role.

To minimize the learning curve for new Gas Controllers and create a seamless transfer of information and expertise, the TC Energy U.S. Natural Gas (USNG) group had been brainstorming solutions.

What if every TC Energy Gas Controller had the opportunity to know what their leadership and predecessors knew from their years of experience on the job?

So, TC Energy set out to mimic these insights…

TC Control Room 2

Brainstorming efficiencies
This led to the opportunity for USNG and Information Services (IS) to collaborate during an Amazon Web Services (AWS) “Working Backwards” session.

The original desire was to create a simple dashboard, however this session led participants to uncover that the end result could be bigger and better. Business acumen is fundamental for our Gas Controllers, and this information needs to be real-time and accessible for them.

Taking it a step further, the question was asked that, if this dashboard was programmed to be real-time and include historical information, could it be smart and provide recommendations for our controllers, as well?

3 days of brainstorming later…Meet Autonomous Pipeline
In collaboration with IS, the AWS session led to the discovery that, by leveraging data from existing systems, and through the application of machine learning – operational outcomes can be predicted. Pulling together data from a variety of sources and evaluating millions of potential scenarios, computer models produce a one-pager of operational recommendations that the Gas Controllers can look to in real-time to assess how best to optimize the pipeline.

Although not fully autonomous, Autonomous Pipeline mimics leadership experience 24/7, and the alerts it provides trigger advanced responses to market conditions.

Early benefits
Today, with Autonomous Pipeline, TC Energy is able to capture additional revenue through the identification of new market opportunities. In addition, we gain efficiencies in equipment run-time, leading to fuel savings, lower greenhouse gas emissions and confidence in reliable and safe operations.

Stay tuned for more updates on this innovation!