Space is a big place, where length is measured by the distance traveled by light over a familiar time interval, such as a light-year. Large distances offer a particular challenge to space exploration; information travels through space at light speed, such that a two-way conversation, like an exchange of hellos, with someone a light-year away would take, literally, two years. This delay — or latency — is a huge handicap for exploration of faraway places.

The two-way latency between Earth and Mars ranges between eight and 40 minutes. The remarkable Mars Exploration Rovers, Spirit and Opportunity, have been commanded by scientists and engineers on Earth who struggle with such latency. Every command must be careful and deliberate because there is no opportunity for split-second changes of plan in the face of unanticipated hazards or discoveries. Spirit and Opportunity have given humans a glimpse of what it might be like to explore Mars in person. Latency will never let martian exploration from Earth using robotic extensions of ourselves be quite like “being there.” We can’t telerobotically extend ourselves faster than the speed of light. It’s the law.

How much latency is too much? From Earth to Moon, the two-way latency is only 2.6 seconds. That made voice communications with the Apollo astronauts relatively straightforward. However, for telerobotic exploration, even that delay can strongly degrade operational efficiency. Our goal for maximum safety and efficiency should be based on our own reaction time, sometimes called the cognitive time scale for a human being. Following research done by cognitive psychologists, and common experience by online gamers, this time scale is a few hundred milliseconds. Networking delays allow workable latencies of 500 milliseconds or less on Earth, which enables telerobotic mining, deep undersea cable and wellhead telerobotic operations, transcontinental telerobotic surgery, and military drone command in a highly cognitive way. We call that cognitive coupling “telepresence.” Half of this two-way cognitive time scale multiplied by the speed of light defines a “cognitive horizon” of 75,000 kilometers, beyond which high-quality telepresence just isn’t achievable.

At present, we do not have a clear plan for human exploration beyond low Earth orbit, in large part because the costs and hazards of putting humans on planetary surfaces are so high. Telerobotic exploration is hobbled by high operational latencies if commands are sent from Earth. The present state of research suggests that we are a long way from replacing the cognitive capabilities of astronaut explorers with autonomous robots, especially in the more complicated environments, which are usually the most interesting ones.

We suggest an alternate approach to telerobotic exploration. We can reduce communication distances to less than the cognitive horizon of 75,000 kilometers with human-staffed teleoperation facilities on spacecraft stationed near exploration targets. In this way, we can precisely control telerobotic surface science and engineering endeavors without sending humans all the way into deep gravity wells. Low-latency teleoperations also have value for exploration of low-gravity bodies such as near-Earth objects.

In many respects, low-latency telerobotics provides operators with heightened awareness and dexterity compared with an astronaut explorer in a space suit. You probably wouldn’t want to do delicate surgery in a space suit, but many surgical procedures have been revolutionized by telerobotics. While many could argue that telepresence will never replace “boots on the ground,” it can be a powerful near-term approach to scientific research and development on other worlds.

On-orbit telerobotic control was proposed many years ago for Mars, and has been recently reassessed by the Human Exploration using Real-time Robotic Operations team at NASA’s Glenn Research Center in Ohio. This plan would send humans into Mars orbit, perhaps onto Phobos or Deimos, where they would control telerobots across the martian surface with the goal of functional telepresence.

In anticipation of future telepresence on Mars or an asteroid, lunar telepresence can be an attractive testbed simply because of its proximity to Earth. For complex and/or delicate lunar surface operations, telerobotics with much lower latency than achievable from Earth, giving real cognitive telepresence, could be highly advantageous. Low lunar orbit could provide that low latency, but an operations post there has serious disadvantages in terms of orbit stability, fragmented communication as target sites quickly rise and set, and frequent shadowing. A highly elliptical lunar Molniya orbit addresses some of those issues, but a telerobotic control node in a habitat at Earth-Moon Lagrange points L1 or L2 would be an even more desirable option.

These Earth-Moon Lagrange points, where gravity is balanced, have been envisaged as important destinations for our travels beyond low Earth orbit. They require little station-keeping propellant, are largely unshadowed and offer uninterrupted lines of sight to the lunar surface, on either the near- or far-side. They are about 50,000 kilometers from the surface, well within its “cognitive horizon.” From such locations, telepresence can be achieved simultaneously across an entire lunar hemisphere, and uninterrupted line-of-sight communication from the teleoperations post to Earth is assured. Such Lagrange point facilities have been proposed as depots for lunar exploration materiel, perhaps even for in-situ resource utilization products. The low energy trajectories from these locations to other Lagrange points in the solar system make a facility there a credible “gateway to the solar system” as well as a job site for servicing the many science instruments that will operate at Earth-sun L1 and L2.

A habitation facility at Earth-Moon L1 or L2 could derive from the technical sophistication gained from the international space station, and could even be constructed and verified at the station before being deployed. While such a facility could have many uses, and be designed in an extensible manner, on-orbit telerobotic control imposes few major architectural requirements. Such control could easily be one of the first goals achieved with such a habitat. A control station there could be responsible for high-dexterity detail work on science instruments, in-situ resource utilization test system management, and even site preparation for future human habitation.

We urge NASA to take advantage of opportunities for on-orbit telerobotic control and telepresence, to advance our exploration goals. With regard to human-robot partnership, such opportunities are quite different from those in which distant robots are operated from Earth, or robots that are working alongside astronauts in gravity wells. On-orbit telerobotic exploration is stunningly extensible. Once we prove it at the Moon, operations at Mars or at vastly less human-hospitable destinations — perhaps the blazingly hot surface of Venus, or the frigid depths of the methane lakes of Titan — are credible. On-orbit telepresence expands the number of potential destinations for human space exploration. It capitalizes on a “being there” kind of exploration that is less about footprints and more about cognitive imprints.

Dan Lester is an astronomer at the University of Texas working with NASA on cis-lunar operations on behalf of science and exploration. Kip Hodges is director of the School of Earth and Space Exploration at Arizona State University and collaborates widely on the development of new protocols for advanced planetary field geology by humans and robots. Michael Raftery is deputy program manager of the international space station at Boeing and has been working with NASA and industry partners on how the space station can be used to support exploration missions.