What an ALifer Has Been Doing About COVID-19
- 1. What an ALifer Has Been Doing About COVID-19 HIROKI SAYAMA [email protected]
- 3. Apex Modeling in Early April (in Response to County’s Request) 3
- 4. 4
- 5. 5 (University President’s statement, April 20, 2020)
- 6. 6
- 7. Projects 1. Campus traffic modeling 2. Campus epidemic modeling 3. New York State county-level visualization 4. Nationwide, worldwide visualization 5. SUNY COVID seed funding 6. Collaboration with local healthcare provider 7
- 8. 1. Campus Traffic Modeling ▪ Agent-based model on multilayer transportation networks ▪ Python + NetworkX + Jupyter ▪ Car road layer + pedestrian path layer ▪ Designated locations (dorms, classrooms, food places, parking lots, bus stops, campus entrances) ▪ 17k pedestrians + 10k vehicles, each with unique behavior reconstructed from data ▪ Objective: Visualize on-campus traffic and measure frequencies and locations of close contacts on a typical Tuesday in Fall 2019 8
- 9. Real Data Incorporated 9 Course schedules and locations in Fall 2019 (from Michelle Ponczek) Individual students’ residence hall and class registration for Fall 2019 (from Michelle Ponczek) Individual employees’ office building and FTE (from Michelle Ponczek) Campus road/path networks (extracted from Google Earth) Square-foot areas of classrooms (from Michelle Ponczek) Parking lot capacities (from Brian Rose) Bus arrival/ departure frequencies (from Brian Rose) Used to develop individual agents’ detailed behaviors Used to count close contacts
- 10. “A Day at Binghamton” 10
- 11. Hypothetical Agent Behaviors 11 Has an actual schedule of Tuesday classes On-campus residents: all transportation on foot (congestion can slow down walking speed down to 50%) Off-campus residents: Commuting by car (need parking) or by bus; after getting off car/bus, all transportation on foot Moves from class to class according to schedule; may go back to dorm if there is enough time Tries a quick trip for food at an appropriate time (if possible) Goes back home when everything is done
- 12. Counting Indoor Contacts N: Number of people in space A: Area of space a: Area of a 3-foot-radius disc Number of close contacts in space = Number of neighbors each person has x number of people / 2 nCC = (N / (A/a) – 1) N / 2 12 (if this is greater than 1)
- 14. 50% Density 14
- 15. 25% Density 15
- 16. Effects of Lowering Campus Population Density 16 R0 = d n pi n = 5 makes R0 = 1 (with d = 4, pi = 0.05)
- 17. 17 2. Campus Epidemic Modeling
- 18. 3. NYS County-Level Visualization 18
- 19. 4. Nationwide, Worldwide Visualization 19 https://p.dw.com/p/3jH59 https://bit.ly/3kwfglo
- 20. 20 5. SUNY COVID Seed Funding Project
- 21. 6. Collaboration with Local Healthcare Provider (in Planning Stage) United Health Services (UHS) Behavioral/epidemiological hybrid modeling of senior living facilities 21
- 22. Lessons Learned • Not what ALife can do, but what we can do (we do have skills!) • Science and theories are there, but real situations are here • We (scientists) know nothing; we must listen, learn, collaborate • Particularly listen to and learn from professionals in the frontline; they are more expert than “experts” • Constraints, constraints, constraints • Things keep changing in a matter of days • People need to make decisions, no matter what 22
- 23. Advantages of Being ALifers • We have technical skills of computation, simulation, visualization • We study complex interactions among heterogeneous agents • We go across multiple scales (micro -- meso -- macro) • We emphasize the importance of space and time • We are exposed to a wide variety of topics, capable of learning more • We know struggle and value of interdisciplinary efforts • We care details and specifics, not just generality or universality • No point in getting in fancy journals; let’s save people’s lives instead 23
- 24. A Moment of Pride I haven’t published a single paper (or posted a single preprint) about COVID-19 this year ◦ Everything was done genuinely to help people ◦ Not to produce universal knowledge, but to derive solutions to very specific local problems for my own communities 24