Why We Need Humans—Not Just Robots—On the Moon

Machines alone won’t be able to satisfy our curiosity.

A nearly full Earth rises above the lunar surface, soon to be visited again by American astronauts.

As luck would have it, I was out of the country—giving a talk at the European Space Research and Technology Centre (ESTEC) on The Value of the Moon. When I arrived back at the hotel, news had already broken that President Trump had re-established the national goal of sending people to the Moon for “exploration and utilization.” My mailbox overflowed with notes from colleagues and science writers. The announcement was hardly a surprise, as at the first meeting of the re-constituted National Space Council in September, Vice-President Pence had said that lunar return was the logical next step for NASA’s human spaceflight program.

Nonetheless, the issuance of the new directive is rightly treated as big news for the space program. And many are asking what I think about it. I approve of it; dropping the Moon as a goal for humans was a major mistake, and now that mistake has been rectified.

We go to the Moon for many reasons, but one of the most important is to learn how to live and work on another world. Although we landed on the Moon six times almost 50 years ago, no one currently at the space agency had any direct contact with that experience. In fact, for the workers of NASA, a lunar return is a wholly new challenge, one informed by what an earlier generation accomplished, but for them, a new task with a steep learning curve and numerous possible pitfalls along the way.

Sixty years on, many still question why humans must go, believing that exploration and science can be accomplished better and more cheaply using robotic explorers. Since human spaceflight is complex, expensive and dangerous, why not let unmanned probes provide all the answers we need? With continuous technical advances in the field of robotics and artificial intelligence, some claim our robotic surrogates in space will not only exceed our physical abilities (which they mostly do already) but our mental ones as well.

One argument against human spaceflight springs from the idea that data collection is the primary function of space explorers, whether that data consists of streams of numbers, images, or rocks collected. Robotic machines can be built that have sensory capabilities that greatly exceed human powers, including measurements that have high precision and lie outside the ranges of human perception. Certainly a robot can be built that could collect ten thousand rock samples per day. And that same machine could measure spectral properties, physical conditions and provide a high definition visual model of any exotic planetary surface to which it was sent. It is claimed that the “productivity” of such a hypothetical mission would greatly exceed anything that a single human could have accomplished. But how is productivity defined?

Data and understanding is not the same thing. When we conduct a mission to explore some aspect of the planets, we seek to understand both process (i.e., how certain properties came to be) and history (i.e., the how and when of these processes, and their sequence in the generation of planetary features). To some degree, scientists on Earth studying the data returned from robotic probes can generate such knowledge. But such knowledge is always incomplete and fragmentary, sometimes to the extent and hazard of being wholly misleading.

Before the Apollo missions went to the Moon, many people believed that the dark, smooth maria were composed of volcanic lava, probably the iron- and magnesium-rich terrestrial lava called basalt. This interpretation was based on images taken from Earth telescopes and robotic spacecraft that showed volcanic landforms, flow fronts and small, cone-like features. Additionally, the robotic Surveyor 5 lander measured the chemistry of Mare Tranquillitatis (the future landing site of the Apollo 11 mission) and found it to be very similar to terrestrial basalt, although with an unusual and unexplained enrichment in titanium.

Believers in the “cold Moon” theory of lunar evolution (most notably, Harold Urey, the Nobel Prize-winning chemist) discounted the evidence of lunar basaltic volcanism. Even after the return of samples from the Moon by Apollo 11, Urey was convinced that samples of basaltic lava in the collection were shock melts produced by impact into the cold accreted dust that he thought made up the maria. Ideas about cold accretion persisted long after the Apollo 11 mission, with both Urey and Cornell astronomer Thomas Gold advocating a “cold Moon” model that they said was validated by the samples (among other things, Gold predicted that there was no bedrock on the Moon). During the Apollo 11 moonwalk, Neil Armstrong (one of the best geology students in the astronaut corps) walked back to the crater he’d flown over before landing and took pictures of the bedrock in Little West Crater, documenting the nature of the mare landing site and confirming the model that geologists favored.

Eventually, Urey admitted that the early Moon had enough internal heat to generate lava (Gold never did own up to it). This seems like an arcane academic argument, but my point is that robotic data alone could not resolve the basic argument. It took not only the samples returned by the Apollo astronauts, but their images and geologic field observations to produce a comprehensive model of understanding. Along with field work, geologists use remote sensing, aerial (orbital) photography and sampling to map the Earth, but no terrestrial geologist would rely solely on that remotely collected data to make a prediction about where to sink a mine shaft or an oil well without personally examining the local terrain and geology.

I speak of geology because that is my field and the one I best understand. Many scientists (particularly those in the “natural sciences” such as geology and biology) will attest to the need for personal interaction with the phenomena they study. We have 200 years of experience with this type of exploration on Earth, and we know what leads to better understanding and what does not. In my own field, as the cost of high-power computing declines, computer modeling of natural processes has become increasingly popular, giving more and more students and workers access to these machines and their more powerful programming tools.

It may well be that a similar effect will be seen when the cost of human exploration drops. And we may be on the cusp of such a revolution. Human spaceflight is expensive because rockets can barely hurl things beyond Earth’s deep gravity well (the “tyranny of the rocket equation”) and we must carry our sustaining environment (life-support) with us. While the advent of lower expense through reusable rockets is touted as a major factor for lowered cost, the need to lug up tons of air, water and other consumables still remains. What changes, then, when we can get those supplies from a nearby local source, one already in space?

I have detailed in many previous posts the Moon’s richness as a source of materials and energy in space. This is (or should be) a primary motivator for human return to the Moon—to use its resources to create new spaceflight capabilities and for life support. While human space travel will never be dirt cheap, we can use cheap dirt on the Moon to lower its costs dramatically. A fueled rocket is more than 90 percent propellant by weight—why not get that propellant from a source already in orbit around Earth? Certainly, many have their eyes set on places beyond the Moon; so learning how to access and use resources on the Moon is beneficial to their getting where they want to go, and vital to remaining there.

People have a value in space beyond the calculus of dollars per kilogram or gigabits per second. We’re told about the accomplishments of the Mars Exploration Rovers, yet, for all the data they’ve collected, we still cannot draw a simple geologic cross-section of their landing sites, and we still do not know the origin of many of the rocks at the site (igneous or sedimentary). A human geologist would have obtained this information after a few hours of fieldwork. People require abundant mass and power, but people give a big return on that investment.

People go into space, like they have everywhere else, because they can and must. Through human creativity and technology, spaceflight will be significantly improved, allowing more people to travel farther and longer than ever before. We must go because our species’ insatiable curiosity demands that we understand our origins, evolution and surroundings, as these factors inform us about our fate and help us influence our destiny—we are intimately tied to the vast universe around us. The history of life on Earth involves death and rebirth; species expire and are replaced by others. In terms of the age of the Solar System, humans have been on Earth for a mere blink of an eye. From studies of the Moon, we know giant impacts periodically destroy life forms on Earth. Perhaps someday, humans will be killed off too, but our descendants living on the Moon and beyond will survive and prosper.

The Moon holds many clues about the evolution and history of our Earth; it provides us with many opportunities to learn how to live and create on another world. With President Trump’s Directive calling for the return of humans to the lunar surface to use the Moon’s resources, we now have the opportunity and national commitment to continue our grand human journey.

Get the latest stories in your inbox every weekday.