Running the collaboration when nobody can travel

Most international collaborations imagine themselves as a series of visits: someone flies out, spends two weeks, sets things up, and flies back. In practice, visits are rare. Budgets get cut, a visa is refused, a term runs long, someone has caring responsibilities.

So the realistic question is not how to make the visits work. It is how the project runs during the eleven months of the year when nobody travels.

What actually breaks

Four things, and only the first is obvious:

Tacit knowledge transfer. Protocols, instrument quirks, judgement calls about ambiguous data — these move easily when two people stand at the same bench and very poorly through written instructions. This is the real cost of not travelling.

Informal problem-raising. In a shared corridor, "this looks odd" gets said in passing. Across a distance, saying it requires composing a message and deciding it is worth someone's attention — so small problems stay unsaid until they are large ones.

Shared sense of pace. Without visible activity, each side assumes the other is either racing ahead or has gone quiet. Both assumptions produce bad decisions.

Relationship among junior members. Senior people already know each other. Doctoral students and postdocs on the two sides often never form a working relationship at all, which is a loss for both the project and for them.

The time zone arithmetic

Vietnam sits several hours ahead of Europe and roughly half a day from North America. That produces two quite different situations, and treating them the same is a common mistake.

With Europe, the overlap is real — Vietnamese afternoon meets European morning. A regular meeting in that window costs nobody their evening.

With the Americas, there is no comfortable window. Someone is up early or late, every time.

The practice that matters here is simple and often skipped: alternate who takes the inconvenient hour. A fixed meeting time that is always 7am for one side and always 8pm for the other reads, over two years, as a statement about whose time counts.

And keep one thing in mind about the calendar as well: national holidays differ, and the Lunar New Year period in particular is a multi-week slowdown that does not appear on a European or American calendar. Plan deadlines around it rather than discovering it.

What survives remote work and what does not

Survives well: analysis, writing, code, literature work, figure preparation, planning, anything with a clear artefact that can be handed over.

Survives with effort: experimental protocol transfer, if you invest in video of the actual procedure rather than a written protocol. Someone filming themselves doing the thing, with commentary, is worth more than any document — and it is a small cost.

Does not survive: training that requires hands-on correction, joint troubleshooting of a physical setup, and the early exploratory phase where a project's direction is still being found. If you need one of these, it is worth spending the whole travel budget on one visit rather than spreading it thin.

Practices that keep a distributed project moving

A short written update on a fixed day, from both sides, whether or not there is news. "No progress this week, teaching load" is a useful message; silence is not. This single habit does more for shared pace than any amount of meeting.

One place where the current state lives — not scattered across email threads. Which documents are current, what is being waited on, what the next milestone is.

Meetings with a written outcome. Five lines after each call: decided, open, who does what by when. Across a language difference, this catches misunderstandings that a call will not — people agree verbally to things they have understood differently.

Deliberate contact between junior members. A separate short call among the students and postdocs, without the senior people, makes a real difference and almost never happens unless someone sets it up.

Language and communication

Two points, both practical rather than sensitive:

Written English is easier than spoken English for most non-native speakers, and video calls at speed are the hardest mode of all. Sending an agenda beforehand and following up in writing costs little and changes how much a call achieves.

Agreement is not always agreement. In a fast call, in a second language, with a senior person on the other side, "yes" often means "I have heard you" rather than "I agree this is feasible." The fix is not to doubt people — it is to ask for the plan back: "how would you approach that, and what would it take?"

When a visit does become possible

Spend it on the things that only work in person. That means: hands-on training, troubleshooting the physical setup, and unstructured time.

The last one is worth defending against a packed schedule. Meals, travel between sites, and unplanned conversations are where the tacit knowledge actually moves, and a visit programmed hour by hour with formal meetings tends to produce less than one with gaps in it.