Most Mars analogue missions come apart long before anyone zips up a simulated suit. Not with a bang — with a missing permit, a rover battery that dies at four in the afternoon, or a budget line marked "miscellaneous" quietly doing the work of three others. UK teams run genuinely useful analogue work in quarries, caves, mines, saltmarshes and upland valleys. The failures are rarely exotic. They are ordinary planning mistakes, repeated every season — and almost all of them are visible months in advance to anyone willing to look hard at their own plan.

What follows is not a list of things that can go wrong. It is a list of things that usually do, and how to plan around them without turning your mission into an administration exercise.

Site selection: fidelity matters more than scenery

The photograph is the trap. A dramatic quarry wall or a windswept ridge looks like Mars in a way that a flat field never will, and that is how teams end up choosing a site that cannot answer their research question. A site can be beautiful, remote and completely wrong.

Start with the science. If you are testing a sampling protocol for fine-grained sedimentary material, a hard basalt quarry gives you the wrong grain size and the lessons you bring home will be thin. If you are studying crew behaviour under a communications delay, the rock matters far less than isolation, cramped living space and the reliability of your link. Decide early which kind of fidelity you are actually buying — geological, operational or psychological — and accept that most UK sites will give you two out of three at best.

Then work through the unglamorous questions, in writing, before you fall in love with a location:

• Can a van, trailer and generator reach camp without a recovery mission of their own?
• What does the site look like after heavy rain, at night, and in February?
• Who owns the land, who manages it, and whose written permission do you need? Protected areas and scheduled monuments usually involve a statutory body, and that can take weeks.
• Are there seasonal restrictions — nesting birds, lambing, deer stalking, footpath closures?
• How far is the nearest hospital with an emergency department, and who is driving?
• Is there mobile signal, and do you want it? A deliberate blackout is powerful, but it needs planning and a real emergency channel.
• Where does the rubbish and the waste water go, and does the landowner know?

Visit in the worst weather you might realistically meet, not the weather on the website. Check who else uses the place, too. A site shared with a paintballing company will not give you the isolation your mission plan assumes. Ask the landowner what has gone wrong for other groups before you — farmers, estate managers and rangers have long memories, and their answers are usually the best free risk assessment you will get.

Crew training: build a research station, not a film set

Crews often train hard on the visible skills — suit procedures, geology, rover driving — and barely at all on what actually breaks missions: fatigue, unclear roles, poor handovers and the reluctance to say "I have had enough".

Treat the crew as the operators of a small, remote research station. They run their own rosters, cook, clean, repair, log data and talk to a support team that cannot see what they can see. Rehearse the boring parts.

• Communications discipline with a realistic one-way delay, including how you phrase a request when the reply is minutes away.
• Shift handovers: what happened, what is unfinished, what is broken.
• Failure drills. Break a connector, disable a radio, run an EVA with a faulty seal — then repeat it after six hours of work.
• Medical and evacuation plans, including who has the authority to call a halt. For anything touching health, fitness or psychological support, take advice from qualified professionals such as your institution's occupational health service rather than relying on goodwill and a shared first-aid certificate.
• One agreed phrase that ends the simulation immediately — "simulation pause", say — and a culture where anyone can use it without apology.
• Conflict and workload: what happens when two people disagree about a call and the support team is thirty minutes away.

Leave room for autonomy. If every minute is scripted from outside, you are testing your schedule, not your crew. Give them a goal, a set of constraints and the authority to reorganise the day when the weather turns — then watch what they do with it, because that is your data.

Train the support team as well. Mission control is where the quiet mistakes live: people talking over each other on the radio link, a shift lead who does not know which experiment is running, a decision that stalls for forty minutes because nobody was sure who was allowed to make it.

Budgets: the costs that surface in week three

Analogue missions are cheap by spaceflight standards and expensive by student-project standards. The classic error is costing the equipment and forgetting everything around it.

• Travel and freight, both directions, including the pallet you did not account for.
• Insurance: equipment, personal accident, public liability. Check what the host site requires.
• Permits, access fees and any monitoring conditions attached to them.
• Consumables that vanish: water, fuel, food, batteries, tape, gloves.
• Spares. Buy two of anything that can break, because it will.
• Data storage, backup, and the time of somebody who knows how to run it.
• Staff and student time, which is real even when unpaid.

Build the schedule backwards from the last day, then add contingency to the parts you cannot move: the permit, the vehicle, the person with the key. When something has to give — and it will — decide in advance what gets cut and who gets to decide.

Write the budget twice. Once as you hope it will go, and once as it will go if the generator fails, the van needs a new clutch and the landowner asks for a deposit. If the second version is unaffordable, the plan is not ready.

Logistics, power and the four o'clock problem

Nearly every analogue mission has a four o'clock problem: the hour when the light goes flat, the crew is tired, and something essential has quietly stopped working. Plan for it deliberately.

Power and charging

• Estimate consumption, then double it, and include the laptops, cameras and phones nobody listed.
• Test every battery in cold, wet conditions before you rely on it. Quarry shade and upland wind are not the same as a warm lab.
• Agree a charging rota in writing. Two people charging from the same circuit at the same time is how you find the fuse box in the dark.

Water, waste and resupply

• Water is heavy and the plan to "refill somewhere" needs a name, a time and a backup.
• Decide what happens to grey water, chemical toilet waste and packaging before you leave, not on the last morning.
• Set a resupply date and treat it as a mission event with its own checklist, not an errand for whoever is free.

Data, documentation and the lessons that survive

The most common waste in analogue work is not equipment failure. It is a well-run mission that produced nothing usable because nobody agreed what a datum was called, what time zone the logs were in, or where the files would live.

• Agree file naming, folder structure and time reference on day one, and write it on the wall.
• Log the failures as carefully as the successes. The broken seal and the abandoned EVA are the most valuable field notes you will bring home.
• Collect the crew's own account while it is fresh — a short daily debrief, recorded, beats a reconstruction three months later.
• Back up nightly, in two places, and make one of them somewhere other than the site.

None of this is glamorous, and none of it will make the photograph. But it is the difference between a mission that looked like Mars for a week and a mission that taught your team something they can use, publish and build on next season.

Plan the mission you can actually run, in the weather you will actually get, with the people you actually have — then spend the saved heroism on the science.

Photo: RDNE Stock project / Pexels