BASALT crews simulating Mars surface work gave researchers a rare clock-by-clock look at how scientific decisions move when every exchange is delayed. In the study, teams working through Mars-like latency used the most text messaging during the middle of an EVA, then saw communication fall sharply once sample collection began. Another number stood out just as clearly: both sides usually needed about 4.6 minutes to notice, read, interpret and compose a reply.
Researchers reported those patterns in Astrobiology while analyzing the BASALT program, short for Biologic Analog Science Associated with Lava Terrains, a field campaign designed to test future Mars operations in volcanic terrain on Earth. The paper, Opportunities and Challenges of Promoting Scientific Dialog throughout Execution of Future Science-Driven Extravehicular Activity, focused on how an Earth-based science team and a Mars-side relay crew actually communicated when the mission moved from scouting to ranking samples and then to collecting them.
Those details matter because a future crew on Mars will not work like astronauts circling Earth a few hundred kilometers overhead. One-way light time between Earth and Mars can range from roughly 4 to 22 minutes, which means many tactical choices in the field will already be moving forward before advice from Earth arrives. BASALT shows that the delay itself is only part of the problem. The human work of turning observations into a useful message also takes time and that extra layer can reshape who truly drives the science during a spacewalk.
BASALT built Mars delay into geology fieldwork
The BASALT project ran high-fidelity analog missions in lava terrains in Idaho and Hawaii, places chosen because volcanic ground can stand in for some of the operational demands of planetary field geology. The program overview in The BASALT Research Program: Designing and Developing Mission Elements in Support of Human Scientific Exploration of Mars lays out that broader architecture, while the crews themselves worked under realistic communication delays of 5 minutes and 15 minutes one way, matching two points within the much broader Earth-to-Mars latency range.
During each simulated EVA, the field astronauts, called EV crew, worked directly at study sites while a Mars-side relay operator called IV2 handled text communication with the Earth-based Science Backroom Team. The teams also used GPS, voice, still images and, in some conditions, video. Yet the paper shows that the crucial scientific dialog across latency happened in the Playbook Mission Log, where messages were queued, delayed automatically and preserved in a written timeline. That wider EVA software backbone is described separately in Future Needs for Science-Driven Geospatial and Temporal Extravehicular Activity Planning and Execution.
Researchers break the EVA into phases because the work itself changes what kind of communication is possible. Early minutes covered translation to the station and broad contextual observations. The middle of the EVA focused on candidate sample search and presampling instrument surveys, stages when Earth had enough time to review incoming evidence and still influence priorities. Later minutes shifted toward actual sample collection, when the crew on the ground had less margin for back-and-forth discussion.
The message surge came before the rocks were bagged
The paper reports that the Earth-based science team sent an average of about 23 text messages per EVA, while IV2 sent an average of about 38. Those messages carried requests, ranking updates, warnings, clarifications and brief explanations of why a sample or observation mattered to the mission goals. Communication was therefore active and sustained, yet it was not evenly spread across the timeline.
Across the analyzed EVAs, message traffic was usually highest between about 50 and 150 minutes into the excursion. That interval matched the candidate sample search and the presampling survey phase, exactly when Earth had the best chance to weigh images, instrument readings and verbal descriptions before committing the crew to a final sampling choice. Once the EVA moved past about minute 200, communication dropped sharply as the field team entered the collection phase and attention shifted from debate toward execution.
The authors are careful with those statistics. They describe the patterns as descriptive rather than as a controlled test proving a single operational law. Weather, distance covered, sample visibility and other mission-specific conditions could shift the curve from one EVA to the next. Even so, the same broad pattern keeps a firm logic: Earth contributes most when there is still time to compare options and its influence narrows once hands-on sampling begins.
Nearly five minutes per reply changed what Earth could do
One of the study’s most useful findings is that delayed science support is slowed by more than orbital physics. On average, IV2 and the Science Backroom Team each used about 4.6 minutes to craft a reply to a direct question or comment. The authors treat that number as a realistic lower limit on the working time needed to notice a message, open it, understand it, decide on a response and type something clear enough to send.
Several details help explain why the number stayed stubbornly high. On Earth, the science team often had to discuss incoming evidence, reach a consensus and then condense that discussion into a short message. On the Mars side, IV2 was not sitting in a quiet office waiting to text. That crew member was following live audio and video, taking notes on field observations, comparing Earth guidance with local expectations and trying to spot misunderstandings before they spread into the EVA timeline.
The paper’s deployment-level numbers show how wide the range could be. Average response time was as low as 2.2 minutes for one Hawaii IV2 set and as high as 7.7 minutes for another, while Earth-based averages ranged from 2.7 to 4.9 minutes. Outliers likely came from overload, delayed notice of an incoming message or a later use of the Mission Log copy feature. For mission planning, the central lesson is simple: once that human response time is added to a 5- or 15-minute one-way delay, Earth cannot manage fast tactical turns the way it can in low Earth orbit.
Audio and video helped, but they also created false confidence
The BASALT paper does more than count messages. It walks through case studies where passive communication misled the Earth team. In one example from the 2016 Hawaii deployment, the Science Backroom Team saw a red patch of alteration and sent a high-priority message asking the crew to place a candidate marker there. Because of latency, the field crew had already marked that exact spot, but IV2 received the instruction without the timing context needed to recognize that Earth was reacting to an older moment. The result was an unnecessary search for a different target and a delay while the teams sorted out what each side thought it had seen.
A second example shows how video can look decisive while still hiding the crucial decision. The EV crew discussed a possible sample site while holding the CB marker and the Science Backroom Team assumed the marker had been placed. In fact, the crew rejected that location because sampling there looked hazardous. Earth, still relying on the audio description and delayed video, ranked the unsampled spot as a top priority before still images could confirm what had really happened.
Those incidents support one of the paper’s strongest operational arguments. Passive observation from Earth can create the feeling of shared awareness without delivering the shared interpretation that science decisions require. A delayed video stream may show a rock, a tool or a gesture, but it does not guarantee that Earth understands which choice the Mars-side crew has actually made. The Mission Log text channel, although slower and less vivid, gave teams a more reliable way to state intent, flag urgency and preserve a record that could be reread when confusion surfaced.
Future Mars EVAs may need more autonomy than Mission Control expects
The recommendations at the end of the paper push toward a different balance of authority than many people imagine when they picture a Mars mission. The authors argue that text-based communication works better than delayed audio because recipients can read it when their immediate tasks allow and can return to it later. They also recommend regular priority updates from Earth, brief rationales for changing guidance and direct attachment of any still image being discussed, ideally with annotations for clarity. NASA’s 2017 Hawaii BASALT field deployment summary describes the same Mars-like 5-minute one-way delay and the Mission Support Center structure that framed those tests in practice.
Another recommendation is more cultural than technical. The scientifically focused IV crew should build its own anticipated priority list from mission goals and field observations, then compare that list against Earth guidance to catch discrepancies early. That practice treats the Mars-side relay crew as an active scientific interpreter rather than as a narrow message courier. The study repeatedly shows that this local judgment helped resolve problems faster than waiting for Earth to reconstruct the scene across delay, a theme explored from the science-team side in Developing Intra-EVA Science Support Team Practices for a Human Mission to Mars.
The paper goes even further by arguing that Earth should resist overmanaging the EVA. Because text response time was about four minutes even before adding one-way light time, short-turn tactical control from Earth becomes reactive almost by definition. BASALT researchers say that reality is one reason the team favored the term Mission Support Center over Mission Control Center. The phrase signals a harder truth about Mars exploration: once astronauts are working in the field with delayed links, the crew nearest the rocks will carry more of the scientific authority.
The study measured an analog, but the warning is hard to ignore
BASALT was an Earth analog, not a flight mission and the authors do not pretend otherwise. Some EVAs were omitted from the aggregate analysis because they did not meet simulation quality criteria and one later EVA followed a unique timeline that would have distorted the comparison. Training levels, staffing and technology in a real Mars program will also improve beyond what an analog field campaign can provide. Even with those limits, the measured patterns remain valuable because they came from people trying to make real scientific choices under realistic operational pressure.
The study also avoids promising that one channel solves every problem. Still images were highly valued, GPS helped with broad situational awareness and audio remained important for the live Mars-side crew. What BASALT changes is the ranking of trust when Earth must influence science at a distance. The fastest-looking channels were sometimes the least reliable for preserving meaning across delay, while the slower written channel often carried the clearest path back from confusion.
For planners thinking about Mars EVA science, the article offers a practical benchmark rather than a cinematic vision. Expect the busiest scientific debate before final sampling starts. Expect each useful reply to cost several minutes of human effort before latency is even counted. Expect video to inform Earth without guaranteeing Earth understands the field decision. And expect the crew on the scene to hold more responsibility, because once the clock passes the middle of an EVA and minute 200 comes into view, Mars will not wait for a perfect answer from home.
