Adaptix is a groundbreaking solution bringing cutting edge artificial intelligence into the pipeline control room. The result of UTSI’s extensive collaboration with Sensewaves, a leading artificial intelligence research and development company based in Paris, France, we offer Adaptix.Pipeline to facilitate real-time business and operational decision making in pipeline control room environments. Adaptix.Pipeline is based on a deep learning AI technology capable of observing and assessing real-time streams of time-series data to perform anomaly detection and predictive analysis, as well as many other analytical functions. Adaptix.Pipeline technology has been successfully implemented across several different industrials, and is proven to provide significant value to critical infrastructure operations and maintenance. Contact UTSI about our pilot project methodology, which is both cost effective and minimally disruptive to ongoing operations.
Implementation of any AI solution is based on a methodology that starts with data, combined with domain knowledge provided by subject matter experts. Through a process of reviewing, cleansing and organizing the data, as well as evaluating various techniques to determine the best approach for attaining the desired outcome, a production quality solution can be produced. While few implementations of AI technology applied to critical infrastructure environments are identical, this approach is generally applicable for most use cases. Further information is available by contacting UTSI.
A few example use cases where Adaptix.Pipeline can provide significant value are briefly described below, although there are many other potential applications for this technology.
In many operating companies, pipeline controllers are obligated to shut down movement of product through a pipeline in the event of a leak alarm that cannot be immediately validated. This can result in a significant revenue impact on the operating company, among other issues.
Adaptix.Pipeline can provide valuable insight to the controller to aid in the validation process by contributing to the controller’s situational awareness, leading to determination of the alarm being a true positive or false positive. If the alarm is determined to be a false positive (invalid), the pipeline can continue product movements without any interruption in service. Adaptix.Pipeline is not a replacement for a traditional leak detection solution, but rather it is a supporting tool that has the ability to correlate current conditions with past history and experience, thus providing the pipeline controller with a much broader understanding of the operating state of the pipeline. This insight can have a direct bearing on the evaluation and ultimate determination of actions to be taken for any given leak alarm.
Pipeline assets periodically require maintenance or they slowly decline to the point of failure. This slow decline is not easily identifiable without constant monitoring of low-profile incoming data. When the asset eventually becomes non-functional, emergency maintenance and loss of productivity can cost pipeline operators significant revenue.
Adaptix.Pipeline constantly monitors these data streams and learns normal operating parameters, and quickly identifies when data streams reflect a periodic or anomalous decline in asset operation. This allows pipeline operators to address the maintenance of the asset before it has caused a significant loss in productivity
Damage from third parties working along a pipeline Right-of-Way (ROW) is the single most common cause of mechanical failures in pipeline infrastructures, and thus, accounts for a significant portion of pipeline leaks every year. One of the major issues is attributed to contractors working along the ROW without first notifying the operator. As work progresses, if a backhoe or other piece of excavation machinery impacts the pipeline, it causes damage, which may or may not result in an immediate failure. Many times such damage occurs many months, or even years, before the pipeline experiences a failure at the point of such impacts, costing the operator significant revenue and downtime to repair.
Adaptix.Pipeline monitors data streams acquired from cathodic protection systems and identifies anomalies in the data reported by these systems, enabling the detection of mechanical impacts to the pipeline when they occur. This allows pipeline personnel to immediately address the issue, and void a costly failure event in the future.
Prior to the implementation of any artificial intelligence/machine learning technology, it is vital to evaluate whether or not the technology is capable of accomplishing the identified objective. Data comes in many different forms and each customer’s objectives are unique, so a proof of concept evaluation can be a necessary first step.
UTSI and Sensewaves have developed a tried and true methodology to ensure that the proof of concept evaluation process is straightforward and effective, and minimizes any disruption on the end of the customer. This process not only evaluates whether Adaptix.Pipeline is a viable product to assist control room personnel, but also examines the viability of the customer’s data for AI/machine learning purposes in case another similar solution could be considered in the future.