Henry Berry
As a cash-strapped oil patch wrestles with the competing issues of staffing and funding, more and more companies turn to various forms of automation/artificial intelligence and robotics in order to do more work with fewer people. In that scenario, drones have flown to the forefront as they’ve become more powerful and versatile.
Advances in recording equipment like cameras and sensors have allowed drones to surpass on-foot humans in efficiency and in detail. Common drone tasks include leak detection, site surveys, security/site disturbances, environmental surveys, and an ever-expanding list of answers to the question of, “While it’s up there, can it do…?”
Henry Berry, director at London-based energy developer Tristone Holdings, listed some of the top ways the oil and gas industry uses drones. “Currently within the industry, drones are primarily being used for maintenance, inspection, and surveillance. Surveillance drones enable remote sites to be monitored, mitigating the need for as many [if any] security personnel. Given the high-value assets housed within an oil site setup, this is going to prove invaluable now and into the future.”
One example of drone efficiency involves pipeline monitoring and inspection. Done by ground-based humans, these inspections can take days to cover even a few miles of pipe. “With drones, on the other hand, this can take only minutes to hours. The inspections themselves tend to be through photographic and video imagery captured by the drones, as well as ultrasonic, thermal imagery cameras, and LiDAR sensors, in some cases,” he said.
Before long, technology now under development will let drones do methane detection.
Berry’s particular focus is on how drones can contribute to oilfield safety. On that topic he explains, “Drones can be used in the predictive maintenance of critical infrastructure, as well as regular inspection work, with activities including inspecting pipelines, flare towers, and confined spaces. All of these can prove hazardous to workers if something goes wrong and they’re needed to get up close and personal, as it were.”
Advances in drone technology and accompanying price drops have made drones more accessible to smaller companies, he said. “Although drone technology is now fairly well-established, its high costs mean that it still tends to be the larger vertical internationals that use them most.” Those large companies usually operate drone technology in-house. But, “From what we’ve seen, mid-scale and smaller O&G firms are beginning to adopt drones into their business models, though their operation and data collection tends to be outsourced to specialist drone companies. Said companies will be called in for the odd inspection, rather than the routine surveillance and monitoring that their larger counterparts are implementing.”
Berry added that drone technology is advancing to the point that it is becoming more of a commodity. The technology is becoming more user-friendly to smaller companies that can’t afford a fulltime drone department. He noted, “The introduction of digital parameters such as geofencing, means that smaller oil companies will be able to carry out drone operations themselves, rather than calling in operatives, at least for the more basic of inspections. Put another way, their actual operation is becoming more risk-free.”
Geofencing refers to the concept of placing a digital fence around the area to be inspected. This would remove the need for an operator to be within visual line of site, or VLOS. He said, “It entirely mitigates the risk of user error leading to drone accidents.”
Shifting Focus
For Peter Walper, founder and president of Midland-based Thermal Cam USA, the sudden downturn was at first a curse, then a blessing. Early March saw several clients agree to large scale projects, only to cancel them with the start of the quarantine and the resulting oil price crash. But things quickly changed for the better.
“The calls are different from what they were a year ago,” Walper reported. “I think that’s because, with the layoff and the early retirements, the manpower isn’t there, but they still need the data.”
Midstream companies in particular still need environmental reporting such as Quad O and other data, whether they have the staffing to do it or not. Many prospective clients tell Walper they no longer have anyone on the ground to do inspections, or that they have heard his drone option is cheaper than the previous system of using helicopters for inspections.
Some companies temporarily looked at doing the inspections and reporting in-house, but balked upon discovering how much time and effort it would take, Walper said. Even if a company already had a drone enthusiast on staff to whom they might assign the task, they soon realized that this person would not have deep-enough expertise to justify the expense of buying their own drone. “And if the guy quits, what then?” Walper asked.
This realization led to a number of the positive calls to Thermal Cam in recent months.
Early drone use was confined to simple tasks such as recording heat or methane plumes. Drones were already creating efficiencies by collecting basic site data previously collected by in-person visits to every site. “They were taking the guy off the ground and saying, ‘Hey, mister pumper, fix it, don’t find it.’ You’re wasting your time finding it because we can find it faster, cheaper, and safer than you can—but we can’t fix it” from a drone.”
As technology improves and data requirements grow, demands on drones are also rising. “What we’re finding now is the robotics is moving heavily toward data gathering for advanced analytics and machine learning,” Walper said. “If you take it to the next level and you say, ‘Hey, this software has artificial intelligence, now you get into a situation of predictive maintenance.”
Producers can run drone-collected data through their ML and AI software to inspect conditions recorded leading up to previous failures to find the common denominator—which event or sequence of events normally precedes a certain problem. When that event starts anew, the producer can take preventive action.
As the need for AI information has grown, Thermal Cam has adjusted its data gathering to provide end users information that can be integrated into their own analytics systems. In light of the greater data requirements, “You have to have a pretty sophisticated drone operator to get out there and provide the data for the AI software people.”
One area many drone operators are looking to is flight rules. Currently the FAA limits drone pilots to the previously-mentioned virtual line of site operation or VLOS. Walper said there are practical issues with expanding that rule. “It’s not so much visual line of sight, it’s how far can the drone fly on what kind of battery power? And number two is how far can you transmit that data back to a central control station, where it’s effectively being sent back to your controller?” The quality of data being transmitted is, for him, the practical limitation.
For longer flights such as pipeline right-of-way surveys, the chase vehicle can keep up with the drone, with the pilot only stopping as needed for battery changes.
Are Unpiloted Drones the Next Wave?
Percepto, an Israel-based technology company with operations in Texas and worldwide, is pushing drone frontiers even further with autonomous “drone in a box” options that do not require the same level of on-site human piloting. Illy Gruber, the firm’s vice president of marketing, said there are ways to work within the VLOS rules to allow the precision and availability provided by autonomous flying. “It’s not always the case that remote operation is allowed from a regulation standpoint—it requires getting a waiver for beyond visual line of sight.” When a waiver is delayed or is not possible, autonomous operation allows one pilot to operate multiple drones.
Drone-in-a-box is largely used to augment security cameras in 24/7 remote operations, Gruber said. The drone can fly on a preset schedule, or when triggered by unexplained motion on the site, or a combination. Between flights it parks itself in its box, which contains a charging pad.
The concept and the company came about in a rather unlikely fashion. On a ski trip the founders, who were hardware engineers, wanted to have a drone follow and make a video recording of them on the way downhill. It needed to be autonomous—using computer vision to allow tracking of an object—because everyone interested in the video would be skiing. When they found nothing like that on the market, the group went to work on development, with the goal of marketing their drone system to consumers. At that time they were mounting their own hardware on existing drones.
Within two years they found greater interest from the security market than from consumers, so they further developed the drone’s capabilities to meet that sector’s greater demands. It was there, said Gruber, that they realized the need to develop the drone-in-a-box concept, which allowed autonomous landing. Now they manufacture the complete package in-house.
Industry, including mining and oil and gas companies, also began to show interest because they were already using drones, just in a more limited form.
In addition to its data collection function, Percepto uses AI to accumulate and process data to provide in-depth insights for clients.
For the oil and gas sector, Gruber said the company’s projects include refineries and petrochemical plants, along with pipeline construction. She said, “The interesting thing is, no matter what the facility is like, all of them are using the drone for conducting various different tasks in different areas.”
Security leads the list of drone jobs, but other uses include visual inspections, thermal camera inspections, and compliance issues. The latter involves making sure employees are wearing personal protection equipment (PPEs), practicing tool and vehicle safety, and more. Because the drone accumulates site data daily, its AI programming can “see” when a truck or other equipment is out of place at the end of the day, Gruber explained. All its data goes to the cloud in real time.
The Future
Gruber and Walper both see the March oil price crash as a catalyst for huge moves into automation and outsourcing, moves that may not be significantly undone in the future.
Walper finds many decision makers asking themselves if they can get more work done with software applications or by farming work out to third parties, in order to reduce the impact of the hiring and firing cycle tied to constant booms and busts. “What’s coming to the front now is the data delivery and how you’re going to utilize it across the whole organization rather than just one little area.”
This commitment to outsourcing is, to Walper, an indication of the future, at least in part due to the third party’s versatility. Instead of needing an in-house person who is knowledgeable about all facets of drone use and technology, “You just find the [company] for the work that needs to be done, and who’s available for any type of work, whether it’s a power line, a methane leak, or something else along those lines,” he said.
This is the third major crash since 2000. With each one the industry has reinvented itself, turning to hydraulic fracturing and significant automation, among other things, to survive and grow. This is not only not your father’s oil patch, it may not be your older sibling’s oil patch either.
______________________________________________________________________________________________________
By Paul Wiseman