Getting an autonomous drone offshore isn't the trickiest part. Landing it is.
A drone returning from an onshore mission lands on a stationary landing pad. An autonomous drone running missions on a floating platform faces a vastly different reality, with both the landing surface and the unmanned aircraft moving simultaneously while a remote pilot oversees operations from afar.
It’s a challenge Shell thought worth solving.
Earlier this year, the operator proved it could safely launch and return its drone in a box at its Mars floating production unit in Mississippi Canyon Block 807 in the US Gulf of Mexico (GOM), which has been renamed Gulf of America. Shell also has become the first GOM operator to secure Federal Aviation Administration (FAA) approval to self-authorize beyond-visual-line-of-sight (BVLOS) drone operations from any Shell facility in the Gulf, which opens the door to Shell’s broader use of autonomous drone operations offshore.
Floating platforms in the Gulf may need to shut in repeatedly throughout their lifetimes in response to hurricane activity. Historically, operators have carried out visual safety inspections of platforms via helicopter flybys before returning personnel to those platforms to bring production back online. But an autonomous drone can provide a more detailed visual inspection more quickly, which could result in a faster return to oil production.
Enter Shell’s drone-in-a-box program.
Drone in a Box
One steamy morning in mid-August, Ross Doak, senior robotics engineer at Shell, stood on the rooftop that serves as the company’s drone test facility in Houston and punched a few buttons on his laptop. About 30 ft away, a small ivory container opened up, a drone powered on and ran initialization checks, and seconds later lifted off for a brief survey of the area around the Shell Technology Center in Houston. After a brief mission, the drone returned to its home base and slowly lowered itself back into the box before the container closed.
Minutes later, Doak sent the same drone on a similar flight, but this time from the air-conditioned comfort of a conference room inside the building. With the control laptop linked to a large wall-mounted monitor, everyone in the conference room could follow the drone’s view in real time.
“How much nicer it is to be able to do a drone-box operation from a nice air-conditioned office,” he said.
The drone-in-a-box setup includes a docking station, an external camera to surveil the drone system and facilitate pre-flight checks, and a gas sensor for compliance with electrical codes. Operated from a bespoke laptop, connectivity is via Starlink and the 5G LTE cellular network. The system can launch, capture data, land, and recharge without a pilot on site.
Offshore, where real estate is at a premium and all equipment must prove its value, the system will be installed atop a building “where nothing else can be mounted anyway,” Justin Trussell, senior maintenance team lead with Shell Exploration & Production Company, told JPT.
For the drone-in-a-box effort, Doak said, Shell has been testing drones from Chinese vendor DJI and US-based Skydio.
“We’ve got very high confidence in the DJI system,” which he called the most prevalent system currently in use within Shell. While Shell has less flight-time experience with the Skydio drone, he said, the team is developing trust in it.
Both systems, though, must overcome challenges related to offshore activities, he said. “Operations over water are different from over land in that the sun reflects far more off water than it does over ground. And so, with vision-based systems, glare is a huge issue,” he said.
Repositioning or adjusting the lens for glare can help when it comes to inspection chores, he said, but navigation is a different story. “In the case of glare, the drone may believe incorrectly that there’s an obstacle very close to it, and it might stop or behave erratically. And of course, that erratic behavior is not what we want,” Doak said.
LiDAR assists with navigation and detects obstacles, he said.
Another complicating factor is that by their very nature, floating facilities … move. And the drone pilot, who is remotely overseeing operations from land or a different offshore facility, isn’t experiencing the same motions.
“We have to be able to account for how much facility motion is on at any given time and make sure the pilot has an awareness of that,” he said. “How do we build the right compute into this to sense that we’ve got a lot of acceleration, heave, pitch, roll, lateral movement, and have the pilot be aware that, hey, you’re getting close to the operational limitations of the drone?”
The drone’s home-base box may move significantly while the drone is on its mission, depending on whether it is deployed from a tension-leg platform (TLP) or a floating offshore production, storage, and offloading vessel.
The drone’s three primary means of determining its position are cameras, global positioning system (GPS), and real-time kinematic (RTK), a satellite navigation technique that fine-tunes location accuracy from meters to centimeters and is essential for helping the drone return to its box, he said.
Mars and Beyond
The team testing the drone’s capabilities at the Mars TLP in 2,940-ft water depth ran 44 flights in April.
Despite the floater moving out of position by as much as 16 ft while the drone was in the air, Doak said, “all of them nailed the landing. We were very happy with that, including a night operation.” The night operation carried a higher risk factor because the vision systems don’t work as well in the dark, he added.
Trussell said bringing a new technology like drone in a box to a late-life asset like the Mars TLP offers more runway to a facility that’s already produced over 1 billion bbl of oil since it started production in 1996.
“We see this technology, especially drone in a box, as being able to reinvent an asset that has some age to it,” he said. “Every asset can benefit from something like this.”
Doak said one of the reasons the drone-in-a-box project received internal funding was to boost hurricane resilience. “If you think about the Gulf of America, the defining operational risk is hurricane season. If we can get to scale quickly on this, if we can be the most-resilient operator during hurricane period, that’ll translate to more production volume. So that’s the big carrot for us.”
Trussell said when platforms must be demobilized due to major named storms, it takes a while before regular operations can resume. “Normally, what we have to do now is wait until conditions are right, put a helicopter in the air” to do physical flybys of the platform to determine if it’s safe to return, he said.
But with the drone in a box, as long as the system maintains backup power, a remote pilot could initiate drone takeoff for a quick inspection as soon as the storm passes. “We’d know if we’re ready to go, and we haven’t had anybody step foot toward a helicopter yet. That allows us ultimately to get back to production sooner because now I know I’m ready to remobilize the crews earlier than I would have,” Trussell said. “If we gain back a day of production following demobilization for a named storm, you’re talking about millions of dollars in value.”
Other Use Cases
Trussell sees other potential uses for the system.
“When I saw the autonomous drone technology amongst some of the other things that we’ve been pitching around, to me, that was one of the most practical pieces of technology I could immediately see where the value would be for us, both in the space of executing inspections more nimbly, as well as some of the HSE cases for the assets with being able to respond to situations without putting people in harm’s way,” he said.
Routine inspections around and above the unit can be done autonomously using the drone-in-a-box system, eliminating the need to bring a drone crew onboard, which can cost as much as $20,000 per trip. He added that just a small number of autonomous inspections could offset the cost of the drone by avoiding crew mobilization expenses.
Jacob Schexnayder, digital transformation lead for Shell's Gulf of America business, told JPT that when Shell experiments with new technologies, new use opportunities can appear.
Because the company already uses drones for inspections, changing the location of the pilot from on-site to a remote location opens up the number of operations that a pilot can support, he said. If there is an issue with simultaneous operations or a weather issue at one platform, the pilot can pivot and oversee drone activity in a different segment of the Gulf, he said.
One potential new use case of the drone in a box is deck space management, he said. The system could take daily pictures looking down from the top of the asset to show the location of materials. “Having that daily snapshot gives us that real operating picture of how to best utilize our cranes when we have a boat coming out” to manage movement of supplies, he said.
Deck surveys related to environmental stewardship are another candidate for autonomous inspection. “We are constantly looking around the water for objects that go overboard and become marine debris, which by law we are required to report to regulators. Here’s now a way that I can do that without tying up a person. Here’s another inspection that we’re doing daily or multiple times a day,” Schexnayder said. “How do we reinvent the way we do the things we have to do to be a responsible operator?”
Right now, he said, it’s not possible to know all the potential use cases for the system, but he suspects many will reveal themselves once the system is available onboard.
While this iteration of drone in a box is not rated for internal operations on the platforms and is instead designed to operate around, above, and below the asset, Schexnayder said a similar opportunity may exist in the future for a caged, properly hazard-rated autonomous drone capable of flying through the facility’s walkways.
Scaling the Program
The April tests at Mars were part of a larger Shell effort to lay the groundwork for a fleet of autonomous docked drone systems based at its floating facilities in the Gulf to support remote offshore infrastructure operations.
Being able to safely launch and land a drone in a box from a floating platform was only part of the picture. Shell needed permission to be able to conduct the drone flights as needed so it would not need to navigate red tape before every mission.
Shell already holds a waiver from the FAA for BVLOS drone operations for all its owned and operated onshore facilities in the US. “That allows us to self-approve because we have a performance-based assurance process,” Doak said.
Shell has now received a similar waiver for BVLOS drone operations across all its owned or operated facilities in the US Gulf. Doak said the waiver, the first of its kind granted by the FAA, allows Shell, under its provisions and application requirements, to self-approve drone operators for BVLOS operations while maintaining compliance with FAA and International Civil Aviation Organization regulations.
“Achieving the waiver a couple months ago was a really big milestone for us,” he said.
The next big milestone is expected by year-end: the permanent installation of a drone-in-a-box system on a floating facility in the Mars Corridor, with both Mars and Olympus targeted in the initial tranche of installations.
Scaling comes next, and Doak expects to add the system to between five and seven Gulf floating facilities by the end of 2027.
Trussell said scaling is a huge part of his expectations for the system over the next 5 years. “The speed that technology moves at, I expect 5 years from now that this same type of technology and concept has expanded to the point that there’s things that we don’t even realize yet that we’re able to do with it,” he said.
And Schexnayder sees the potential for even more autonomy in the future. “Changing out this payload to match the mission requirements … is going to be instrumental in truly unlocking the full autonomy to scale,” he said.
