The multi-hub academic conference

Dynamic, sustainable, inclusive


Richard Parncutt, 2023

Global conferences are the lifeblood of academia. For a few intense days,
experts from all over the world  exchange information and ideas, and plan future projects. They have a great time seeing old friends and doing a bit of tourism. To a large extent, that's how academia progresses in the modern world.

Until recently, it has been normal for hundreds of colleagues in a given academic discipline to fly to a single location, from anywhere in the world. That is increasingly seen as problematic, for two reasons:
Various new conference formats have emerged that address these problems in different ways:  
The global timetable

Global conferences with a virtual, VR, or multi-hub format can bring the world together and include everyone, while at the same time minimizing emissions. They also suffer from an intrinsic timetabling problem. Everyone has to sleep, and people prefer to do that at night. Night happens at different times around the globe. A practical solution has two parts:

First, video-record all content so everyone can view everything later. Tip: YouTube live streams are automatically and immediately available to be viewed as videos.

Second, design the program to maximize real-time interaction; like this:

This configuration allows every hub to communicate in real time with every other hub, every day for either 4 or 8 hours. To my knowledge, no other humanly reasonable solution achieves that level of global connectivity. Further detail is  here. For a general introduction to multi-hub conferences, see these papers (Zechlau et al, 2023Parncutt et al., 2021or this video.

Some tips for organising multi-hub conferences

Hub locations. Before approaching an international colleague with a request to organize a hub, plan in advance the daily timetable at that hub. If participants at that hub will have to get up very early or go to bed very late, it is better to try a different location, in a different time zone. For a detailed solution to this problem, again see this link.

Technology. It is nice to have wireless microphones for discussions after talks. But that is surprisingly difficult to organise and creates a list of things that can go wrong. It also means you need a technician in each conference room. The following simplified solution requires neither technical support nor special equipment. The speaker in each live presentation speaks directly to a laptop, using the internal microphone or a headset in the usual way, like in a regular Zoom call. The laptop screen is projected so the audience can see it, both at the live presentation and at virtual presentations at other locations. The laptop sound goes to an amplifier and loudspeaker at the back of the room, turned down relatively low to avoid acoustic feedback. During the discussion, people with questions come to the front of the room and talk to the laptop -- both in the room where the talk is physically held and in other rooms where the talk is watched remotely, via another laptop. In the remote room, the laptop can point toward the audience during the talk with the microphone turned off so the speaker can see the remote audience. At the end of the talk, the laptop in the remote room can be turned 180° and the microphone turned on as audience members come forward to ask questions. In Zoom, each laptop is a user, and the username is the name of the room, so anyone watching the talk can see where it is happening. For this to work, all rooms need moveable seating so audience members can easily move around during talks. There are many other solutions to this problem, and each hub can solve it differently, depending on available equipment and expertise.

Global plenaries. At a traditional conference, a plenary session is one in which all participants meet in the same large room. In a multi-hub conference, a plenary that includes all hubs is only possible if some colleagues work during the night. For that reason, opening/closing sessions and global society meetings happen twice, in two different (consecutive) 4-hour slots. Participants at the second meeting are informed at the start about happened at the first meeting, and/or have the opportunity to watch a video of the first meeting before the second one starts. The result is more inclusive than a plenary at a traditional single-location conference, which excludes colleagues who cannot afford the total cost of travel, accommodation, and registration. Both formats have advantages and disadvantages.

Finance. Each hub is a mini-conference with its own budget and registration fee. Registration is lower in countries with lower GDP per capita, to encourage inclusion. Expenses for hub organizers include room rent, equipment, refreshments, support, and entertainment. Income may include a government or university grant, so the first thing to do after deciding to create a hub is to look for financial support. After that, hub organizers will estimate what registration fee will be needed to cover costs. If the organizing hub runs a peer-review procedure on behalf of all hubs, the additional expenses can be met by asking all active and passive remote participants (those who do not travel to a hub) to pay their (reduced) registration fee to the organizing hub by credit card. Additional possibilities include grants and transfers from other hubs.

Regional societies.
Each hub creates a new sense of regional academic community among physically present participants. That can help leading local colleagues establish or strengthen regional academic societies -- an aspect that is especially interesting in the Global South, if the conference's research area is less clearly established and institutionalized there than in the Global North.

How many hubs? 

There are many different possibilities for the number and location of hubs. The following is written on the assumption that the conference should have global character, including colleagues from anywhere in the world, and treating them equally while at the same time minimizing emissions.

If we want each hub to be in real-time contact with other hubs for the entire working day (8 hours), and with any given hub for at least 4 hours per day, and if we want to give participants at each hub at least 12 continous hours of rest each night (e.g., 9 pm to 9am), then the minimum number of hubs is 3. In that case, the hubs should be exactly 8 hours apart (e.g., Tokyo, Los Angeles, London). 

A 3-hub conference is more inclusive and sustainable than a single-location event, but it is still not ideal. Regarding inclusion, if the hubs are placed on the richer continents (e.g., North America, Europe, Australia), then participants on poorer continents (South America, Africa, Asia) can't afford the trip. Regarding sustainability, if the hubs are on poorer continents, the richer
participants will fly intercontinental and the carbon footprint of the conference wll be too high.

A promising solution is to place hubs on all 6 inhabited continents. In that case, it's ok if spacing around the globe is uneven, provided the distance between adjacent hubs does not exceed 8 hours. If an extra hub or two are added (e.g., East- and West-coast USA, or India and Korea), there might be 7 or 8 altogether.

A much larger number of hubs is also possible. The biggest academic conferences have over 10,000 participants (e.g. in the USA: Society for Neuroscience, Chemical Society, American Geophysical Union). If these went multi-hub,
they could theoretically be split across 100 hubs with 100 participants each, with a hub in most countries and several in large countries like the US or China (with timetable limitations as explained here). Both the reduction in emissions and the increase in cultural diversity would be enormous. Are bigger conferences waiting for smaller ones to show them the way?

In all these cases, it is possible to give each hub the same rights and status as all others, regardless of size. Each can have two 4-hour working timeslots, one in the morning and one in the evening. Smaller hubs can have fewer rooms and show fewer of the available presentations. Larger hubs can have more rooms and both the live and the virtual program can be bigger. Participants can always be free to watch other presentations privately, either live or as videos. 

The optimal number of hubs thus depends on
two opposing criteria:
An optimal compromise depends on the expected total number of active participants. If that's about 1000, the optimal number of hubs might be 6, with one hub on each continent and >100 participants at each. In the 4-hub conference that I organized in 2018, there were 300 participants at the organizing hub in Austria, 150 in Canada, 100 in Argentina, and 50 in Australia, making 600 altogether. The main problem with that conference was the big time difference between Montreal and Sydney (10 hours). A future conference could solve that problem by having either equally spaced hubs or a larger number of hubs, for example 7 hubs, so as to include all inhabited continents, with separate hubs for South Asia and East Asia. We might then expect 300 participants at the European hub, 100 in South Africa, 100 in India, 100 in Korea, 50 in Australia, 250 in the USA, and 100 in Brazil, plus 100 active remote participants, total: 1100. Of course, these are only very rough estimates.

The administrative load for the organizing a hub can be reduced by making the hubs more independent. That can be done by allowing them to write their own program, or at least the virtual part (following guidelines from the organizing hub), or by asking hubs to solve technical problems together without the organizing hub (for example, two hubs in almost the same time zone can meet virtually).

As at any conference, hub organizers will need contingency plans in case things go wrong. Technical failure is one example. A list of possible problems and trouble-shooting strategies can be helpful. If a hub fails to transmit a talk at an arranged time (because the speaker is ill, or there is a transmission problem), another virtual talk can be added to the programs of other hubs at the last minute. First preference might be given to a real-time transmission from another hub. Otherwise, a video from a previous talk at another hub could be shown. To prepare, every hub could make a list of talks they aren’t showing (usually because they are happening at night), perhaps in order of reviewers' grades. Each talk would have its own Zoom link (or similar, if real-time) or YouTube link (or similar, if not real-time). It would also be possible to leave a gap in the program -- participants might be grateful for extra time.

Seen from that perspective, not much can go wrong that could not go wrong at a traditional conference. The internet is not going to go down, but it is important to check in advance that all rooms in all hubs have a fast internet connection, and to check the speed at different times of day and different times in the year, to be sure. At the conference I organized in 2018, no talk was canceled or even delayed for technical reasons.

The opinions expressed on this page are the authors' personal opinions.
Suggestions for improving or extending the content are welcome at
Back to Richard Parncutt's homepage