• 01.10.13

The Next Mars Invasion Will Be Oil-Drilling Robots That Don’t Cause Spills

Because there’s no way to fix things 140 million miles away, the Mars rovers have to be able to prevent their own accidents as they drill. It’s technology that now is being imported back from space to help ensure safer oil drilling on Earth.

Orson Welles’ apocryphal radio broadcast of a Martian invasion may have been true after all. Retelling the War of the Worlds to an English audience in 1932, the author foretold of hordes of multi-tentacled Martians invading the Earth in search of precious resources. Today, multi-armed robots built by NASA drilling for water and life in Mars may soon be on Earth again. This time, they will be hunting for oil and gas.


Autonomous drilling systems probing the Earth’s crust to extract hydrocarbons has long been a dream of the oil and gas industry. Ever since the first patent for drilling automation was awarded to Shell in 1942, companies have tried to automate the process to cut costs, raise efficacy and improve safety–with limited success. After the environmental disaster of BP’s 2010 Deepwater Horizon oil spill in the Gulf of Mexico, that effort took on new urgency. At fault in that incident, reports the National Academies of Science (PDF), was a string of human errors and equipment failures, many likely preventable by automated systems.

NASA has finally moved the fossil fuel industry within reach of its goal. The space agency’s Martian rovers faced an extreme problem: drilling, sampling, and fixing foul ups happening millions of miles from any human hands. Even traveling at the speed of light, problems detected on the Martian surface would take 10 minutes to reach Earth and then another 10 minutes for instructions to return to Mars.

That wasn’t good enough, said Brian Glass, a NASA scientist leading up the effort to perfect the Mars drilling, “Drills get stuck all the time,” he said in a NASA release. “It’s still an art form for humans drilling on Earth,” Instead, NASA decided to tap three kinds of artificial intelligence–model-based reasoning, neural networks, and table-based, fault matching–along with sensors such as laser vibrometers and accelerometers to detect and fix problems on the rig as they happened. Tests in the lunar-esque landscape of Spain’s Rio Tinto and the frozen wastes of the Arctic tundra were largely successful. The teams logged more than 40 hours of autonomous operation with no human intervention.

Now, the companies that helped NASA with its software and hardware for Mars are designing intelligent drill bits that can detect, respond and resolve problems on their own from anomalous pressure spikes to mechanical jams. Eventually, industry experts predict, “fully automated rigs [will] roll onto a job site using satellite coordinates, erect 14-story-tall steel reinforcements on their own, drill a well, then pack up and move to the next site,” reports Bloomberg.

But they’re not there yet. Energid, Seabed Rig, and Robotic Drilling Systems (RDS) are among the first firms to combine the hardware and software letting oil companies test on real drill sites. RDS’ early series of drilling robots can rise ten feet, swap out 15 hand attachments and and lift more than a ton, taking on repetitive tasks now done by rig pipehandlers and deckhands.

Even so, most of the work still happens by the hands and heads of humans. Experts caution the drive to automate something as dangerous and unpredictable as oil drilling will take decades–if not more–to perfect. Perhaps the next Martian invasion will help.

About the author

Michael is a science journalist and co-founder of Publet: a platform to build digital publications that work on every device with analytics that drive the bottom line. He writes for FastCompany, The Economist, Foreign Policy and others on science, economics, and the environment.