Boost the Safety and Efficiency of Pipeline Maintenance with Unmanned Technology Posted on January 31, 2020 When we think of autonomous, unmanned technologies – our minds can sometimes go to a pessimistic place. Unemployed workers standing out in the streets, while unflinching robots work without a moment’s lapse in the factory. The reality isn’t so horrifying. These technologies can be used to handle the arduous and dangerous tasks that burden your team, freeing them up to make the absolute best of their talents. The fact is machines need humans because, no matter how smart a robot may be, there are always situations where humans and machines must work together. Even the best AI can’t solve the types of problems humans can – yet. In the case of oil & gas pipelines, new technologies like Artificial Intelligence (AI) and Unmanned Aerial Vehicles (UAVs) can be put to work for the greater good of your field technicians, allowing them to work in a safer and smarter fashion. With the high-quality, consistent datasets these technologies yield, you can harness the power of predictive analytics to focus your crew’s efforts on preventive maintenance on specific problematic areas, rather than the entirety of the pipeline itself. Put Unmanned Hardware, Not Your Staff, in Harm’s Way Let’s face it, pipeline inspection and maintenance can be dangerous work. When looking at an 8-year summary of the 5,512 pipeline incidents in the United States from 2010-20181, we can see the following averages in statistics: Every day, 1.7 pipeline incidents are reported, each of which cause approximately $1.3 million in property damage. Every 4 days, a pipeline catches fire. Every 5 days, an injury occurs as a result of a pipeline incident. Every 11 days, an explosion occurs on a pipeline. Every 26 days, a fatality occurs as a result of a pipeline incident. The average human cost of these incidents per year equals approximately 73 injuries and 14 fatalities. So, what can operators do to improve safety and mitigate risks to their staff? Oftentimes, maintenance crews are sent in to verify an anomalous sensor reading from the pipeline SCADA system. A SCADA alert is only an indication of a problem along a vast segment of pipeline, lacking a visual representation and detail on the precise nature and location of the incident. When a ground crew is sent to inspect an affected segment of pipe, they could be heading directly into a volatile situation. While aerial inspection crews may be able to avoid the immediate dangers of certain pipeline incidents, due to their altitude from the immediate hazard area, accidents involving aircraft in asset inspections are not uncommon. There will always be an inherent degree of risk in sending your staff as the first response. With an autonomous UAV, you don’t need to assume this risk. If a UAV suffers a crash-initiating failure or takes damage from volatile pipeline damage during an inspection, it can simply be repaired or replaced. That being said, with considerate mission-planning and thorough maintenance, losing your aerial system won’t be a frequent concern, but the cost of losing an autonomous piece of hardware is incomparably preferential to losing a human life. Focus Your Team on Addressing Problems, Not Looking for Them Collecting raw imagery during an inspection is just the first step in identifying issues. Even with high-resolution photos and specialized sensor imagery, analyzing the sheer volume of images from each flight is a fatigue-inducing job. Thankfully, computer vision software with machine-learning algorithms can greatly expedite the workload of analysis. Quickly scanning through the raw imagery from inspection, the software can flag visual anomalies ranging from obvious issues – such as active spills or third-party activity – to more subtle warning signs like vegetation encroachment. With a picklist of potential anomalies generated from the machine analysis, your team can verify and disqualify actionable issues on your pipeline with confidence. As you continue to give the system feedback from each inspection, the software will learn what warning signs to look for around your asset. Making this kind of aerial data solution a routine part of your inspection not only improves the ability to perform efficient data analysis but also paves the way to predictive analytics models. As issues are recorded, you can leverage your historical data sets to determine problematic segments of your pipeline that need to be addressed. Rather than sending crews out to look for active issues across thousand kilometers of pipeline, you can direct them to a specific segment to perform proactive maintenance. By leveraging unmanned aerial systems and AI for visual inspection and data analysis, you can make the best use of your team’s technical expertise. Rather than spending their valuable time on lengthy inspections or sending them blind to a high-risk situation, your team can focus their efforts on the actions that will have the best results on pipeline integrity. Not to mention, shift their efforts from remediating urgent issues to performing proactive maintenance. A Remote Set of Eyes Wherever and Whenever Your Crew Needs Not only are UAVs a safer option that provides high-quality data, but they also provide much greater operational flexibility for performing visual inspections. A UAV operating with a high level of autonomy is the perfect vehicle for performing routine visual inspections across vast distances of land and in remote locations. When paired with strategically-placed launch and recovery stations for take-off, landing and recharging – the effective operating range of a commercial-grade UAV is near-infinite. With the right infrastructure in place, a single autonomous aircraft can inspect the entirety of your pipeline. For true operational excellence, however, creating a distributed network with multiple aerial vehicles, each with their own xStation, makes it possible to perform missions-on-demand, wherever and whenever you need. SkyX’s xStations provide a home base that the aerial system can autonomously take-off from, land and recharge between missions. This operational flexibility doesn’t come at a detriment to the quality of visual inspection either. In fact, UAVs are able to capture incredible aerial data. Equipped with high-resolution cameras, a UAV can capture imagery with detail down to the inch. Not to mention, when an aerial vehicle has the carrying capacity for multiple payloads, you can use these specialized sensors – such as LiDAR or gas spectrometers – to paint a more holistic picture of the state of the problem. Have questions about how high-quality aerial data can elevate your organization? Contact our team to discuss your unique challenges and data requirements. 1 Pipeline Incidents Continue to Impact Residents, Matt Kelso, Fractracker Alliance, 2018