The multi-hub academic conference: Sustainable and inclusive

The global multi-hub academic conference

Choosing hubs and sketching the program

Richard Parncutt, 2023

A multi-hub conference is a one that is split across several hubs in different continents. For an introduction to this idea, follow this link.

A promising way to create a global conference program that treats all global locations equally is to divide the 24-hour day into three equally spaced 4-hour timeslots (e.g., UTC 0-4, 8-12, and 16-20). The entire conference happens within those slots:

Each day, each hub is active for two 4-hour slots (morning and evening). In the morning, the hub communicates with hubs toward the East, and in the evening toward the West. In the third timeslot (night), the hub is inactive. Next day, participants check out what happened the previous night by watching videos, either individually or in programmed sessions.

The 4-hour break between morning and evening is called siesta. During this time, the program.is empty. Conference participants are encouraged to have a long relaxed lunch and lie down for an hour. No official meetings, please (people can use the 30-min breaks, and organizers can create a separate program timeslot for meetings). After all, there's an exciting evening coming up, followed by another exciting morning, and some will go out for a drink after the evening session.

Each 4-hour working slot is divided into two 2-hour halves, each of which starts with a 30-minute break (for recreation and informal communication) followed by a 90-minute work period. If participants are present for the break (regardless of which activity interests them), the following working session can start exactly on time.

Breaks coincide across hubs. Informal communication is either planned in advance or spontaneous. In planned meetings, consenting participants with similar interests are grouped together.

The 90-minute work periods are divided up in different ways, e.g., three regular 30-minute talk slots, or a 60-minute keynote followed by a 30-minute discussion. If possible, subdivisions coincide across hubs.

Evaluation of program solutions at individual hubs

A global multi-hub conference works well if it includes three roughly equally spaced hubs, 8 hours apart, plus other hubs that are not too far away in time from the initial three. All hubs are nominally equal (no central hub, no satellites). That usually means including hubs on the Pacific rim (e.g., Sydney or Tokyo, and Los Angeles or Vancouver).

That way, a morning slot from 8 am to 12 noon can be shared with a hub toward the East, at which the time will be no later than 4 pm to 8 pm (evening). Similarly, an evening slot from 4 pm to 8 pm can be shared with a hub toward the West, at which the time will be no earlier than 8 am to 12 noon.

At each hub, the start and the end of each working day are 12 hours apart. For example, if the day starts at 9am, it ends at 9pm. The local time of the start and the end depend on the global configuration of hubs.

The "good" hub program. The most comfortable solution at a given hub is a working day that starts and ends at 8 or 9. If the day starts and ends at 8, the siesta is from 12 noon to 4 pm. If the day starts and ends at 9, the siesta is from 1 pm to 5 pm. In the following, I will call these solutions "good": they are not perfect (most people would prefer a shorter siesta), but given the global constraints of the conference they are probably the best possible solutions.

The "ok" hub program. Depending on location and timezone, the working day might run from 7 am to 7 pm, 10 am to 10 pm, or 11 am to 11 pm. In the following, I will call these solutions "ok". They are not great, but people can adjust to them more easily than recovering from jetlag after a typical intercontinental flight. The advantage of communicating in real time with all other hubs outweighs the disadvantage of strange working hours.

The "poor" hub program. It is not a good idea to start at 6 am or earlier at a given hub (participants will need breakfast), or to end at midnight or later. I will call such solutions "poor" and avoid them.

Planning hub locations

Before approaching international colleagues who could organize a hub, an organizer should plan the daily timetable at that hub and ensure that the hub can be realistically included, given other hubs that have already been chosen. The timing at each hub should be "good" or "ok".

The following tables can help organizers locate hubs. The listed hub locations are examples only. They can be replaced by other locations in the same time zones. They cover all 6 populated continents but not all time zones.

If the three 4-hour blocks start at whole numbers of hours after UTC=0, organizers have 8 options to choose from. They are analyzed in the following 8 tables. In each table, the first row shows the working times of the conference relative to UTC (GMT). The other rows show local times at each location on a 24-hour clock.

Clearly, some options work better than others. Option 6 (UTC 6-10, 14-18, 22-2) looks especially promising -- at least for the chosen time zones.

The analysis applies only for the Northern summer (August 2023). Things change slightly for the Northern winter, because there is daylight saving time in some countries but not others. The table has been created using timeanddate.com.

Option 0, with working times UTC 0-4, 8-12, and 16-20

Quick summary of results for this option:

Good hubs: London (=Lisbon), Cape Town (=Berlin), Beijing, Tokyo, Los Angeles
Ok hubs: Tel Aviv (=Helsinki), Melbourne
Poor hubs: Delhi, New York, Sao Paulo,,Honolulu

Possible hub	Time difference	UTC 0-4	UTC 8-12	UTC 16-20	Evaluation
London (=Lisbon)	1	(1-5)	9-13	17-21	good
Cape Town (=Berlin)	2	(2-6)	10-14	18-22	good
Tel Aviv (=Helsinki)	3	(3-7)	11-15	19-23	ok
Delhi	5.5	5-9	13-17	(21-1)	poor
Beijing	8	8-12	16-20	(0-4)	good
Tokyo	9	9-13	17-21	(1-5)	good
Melbourne	10	10-14	18-22	(2-6)	ok
Honolulu	14	14-18	(22-2)	6-10	poor
Los Angeles	17	17-21	(1-5)	9-13	good
New York	20	20-0	(4-8)	12-16	poor
Sao Paolo	21	(21-1)	5-9	13-17	poor

Option 1 (UTC 1-5, etc.)

Good hubs: Beijing
Ok hubs: London, Cape Town, Tokyo, Melbourne, Honolulu, Los Angeles
Poor hubs: New York, Tel Aviv, Delhi, Sao Paulo

Possible hub	Time difference	UTC 1-5	UTC 9-13	UTC 17-21	Evaluation
London	1	(2-6)	10-14	18-22	ok
Cape Town	2	(3-7)	11-15	19-23	ok
Tel Aviv	3	(4-8)	12-16	20-24	poor
Delhi	5.5	6-10	14-18	(22-2)	poor
Beijing	8	9-13	17-21	(1-5)	good
Tokyo	9	10-14	18-22	(2-6)	ok
Melbourne	10	11-15	19-23	(3-7)	ok
Honolulu	14	15-19	(23-3)	7-11	ok
Los Angeles	17	18-22	(2-6)	10-14	ok
New York	20	21-1	(5-9)	13-17	poor
Sao Paolo	21	(22-2)	6-10	14-18	poor

Option 2 (UTC = 2-6, etc.)

Good hubs: Beijing, Honolulu
Ok hubs: London, Delhi, Beijing, Tokyo, Los Angeles, Sao Paolo
Poor hubs: Cape Town, Tel Aviv, Melbourne, New York

Possible hub	Time difference	UTC 2-6	UTC 10-14	UTC 18-22	Evaluation
London	1	(3-7)	11-15	19-23	ok
Cape Town	2	(4-8)	12-16	20-0	poor
Tel Aviv	3	(5-9)	13-17	21-1	poor
Delhi	5.5	7-11	15-19	(23-3)	ok
Beijing	8	10-14	18-22	(2-6)	ok
Tokyo	9	11-15	19-23	(3-7)	ok
Melbourne	10	12-16	20-0	(4-8)	poor
Honolulu	14	16-20	(0-4)	8-12	good
Los Angeles	17	19-23	(3-7)	11-15	ok
New York	20	(22-2)	6-10	14-18	poor
Sao Paolo	21	(23-3)	7-11	15-19	ok

Option 3

Good hubs: Delhi, Honolulu, Sao Paolo
Ok hubs: Beijing, New York
Poor hubs: London, Cape Town, Tel Aviv, Tokyo, Melbourne, Los Angeles

Possible hub	Time difference	UTC 3-7	UTC 11-15	UTC 19-23	Evaluation
London	1	(4-8)	12-16	20-0	poor
Cape Town	2	5-9	13-17	(21-1)	poor
Tel Aviv	3	6-10	14-18	(22-2)	poor
Delhi	5.5	8-12	16-20	(0-4)	good
Beijing	8	11-15	19-23	(3-7)	ok
Tokyo	9	12-16	20-0	(4-8)	poor
Melbourne	10	13-17	(21-1)	5-9	poor
Honolulu	14	17-21	(1-5)	9-13	good
Los Angeles	17	20-0	(4-8)	12-16	poor
New York	20	(23-3)	7-11	15-19	ok
Sao Paolo	21	(0-4)	8-12	16-20	good

Option 4

Good hubs: Delhi, Honolulu, New York, Sao Paolo
Ok hubs: Tel Aviv
Poor hubs: London, Cape Town, Beijing, Tokyo, Melbourne, Los Angeles

Possible hub	Time difference	UTC 4-8	UTC 12-16	UTC 20-0	Evaluation
London	1	5-9	13-17	(21-1)	poor
Cape Town	2	6-10	14-18	(22-2)	poor
Tel Aviv	3	7-11	15-19	(23-3)	ok
Delhi	5.5	9-13	17-21	(1-5)	good
Beijing	8	12-16	20-0	(4-8)	poor
Tokyo	9	13-17	(21-1)	5-9	poor
Melbourne	10	14-18	(22-2)	6-10	poor
Honolulu	14	18-22	(2-6)	10-14	good
Los Angeles	17	(21-1)	5-9	13-17	poor
New York	20	(0-4)	8-12	16-20	good
Sao Paolo	21	(1-5)	9-13	17-21	good

Option 5

Good hubs: Tel Aviv, New York
Ok hubs: Cape Town, Delhi, Melbourne, Honolulu, Sao Paolo
Poor hubs: London, Beijing, Tokyo, Los Angeles

Possible hub	Time difference	UTC 5-9	UTC 13-17	UTC 21-1	Evaluation
London	1	6-10	14-18	(22-2)	poor
Cape Town	2	7-11	15-19	(23-3)	ok
Tel Aviv	3	8-12	16-20	(0-4)	good
Delhi	5.5	10-14	18-22	(2-6)	ok
Beijing	8	13-17	(21-1)	5-9	poor
Tokyo	9	14-18	(22-2)	6-10	poor
Melbourne	10	15-19	(23-3)	7-11	ok
Honolulu	14	19-23	(3-7)	11-15	ok
Los Angeles	17	(22-2)	6-10	14-18	poor
New York	20	(1-5)	9-13	17-21	good
Sao Paolo	21	(2-6)	10-14	18-22	ok

Option 6

Good hubs: Cape Town, Tel Aviv, Melbourne
Ok hubs: London, Delhi, Tokyo, Los Angeles, New York, Sao Paolo
Poor hubs: Beijing, Honolulu

Possible hub	Time difference	UTC 6-10	UTC 14-18	UTC 22-2	Evaluation
London	1	7-11	15-19	(23-3)	ok
Cape Town	2	8-12	16-20	(0-4)	good
Tel Aviv	3	9-13	17-21	(1-5)	good
Delhi	5.5	11-15	19-23	(3-7)	ok
Beijing	8	14-18	(22-2)	6-10	poor
Tokyo	9	15-19	(23-3)	7-11	ok
Melbourne	10	16-20	(0-4)	8-12	good
Honolulu	14	20-0	(4-8)	12-16	poor
Los Angeles	17	(23-3)	7-11	15-19	ok
New York	20	(2-6)	10-14	18-22	ok
Sao Paolo	21	(3-7)	11-15	19-23	ok

Option 7

Good hubs: London, Cape Town, Tokyo, Melbourne, Los Angeles
Ok hubs: Tel Aviv, Beijing, New York
Poor hubs: Delhi, Honolulu, Sao Paolo

Possible hub	TIme difference	UTC 23-3	UTC 7-11	UTC 15-19	Evaluation
London	1	(0-4)	8-12	16-20	good
Cape Town	2	(1-7)	9-13	17-21	good
Tel Aviv	3	(2-8)	10-14	18-22	ok
Delhi	5.5	(4-8)	12-16	20-0	poor
Beijing	8	7-11	15-19	(23-3)	ok
Tokyo	9	8-12	16-20	(0-4)	good
Melbourne	10	9-13	17-21	(1-5)	good
Honolulu	14	13-17	(21-1)	5-9	poor
Los Angeles	17	16-20	(0-4)	8-12	good
New York	20	19-23	(3-7)	11-15	ok
Sao Paolo	21	20-0	(4-8)	12-16	poor

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