Graduate School of Business Sciences, University of Tsukuba
Model Specifications
1. Visualization of contact infection risks at home after returning home from going out
This model recreates the state in which a virus that has adhered to one’s hands and belongings when going out later propagates to various items in one’s home. It is also possible to visualize the risk of infection in the case of family members undergoing home treatment.
2. Data used
Data were extracted on location, behavior, and contact items from approximately 100 post-return home behavior surveys conducted by Lion Corporation, and various tables were created including behavior history tables, behavior probability tables, contact probability tables, visit frequency tables, contact frequency tables, and stop condition tables.
Table of material transmission rates were created for virus laboratory experiments, measuring the rate of virus transmission after droplet contact.
3. Agent-based model for visualization of contact infection risks at home
With the automatic layout creation function, multiple floor plans are set up, and based on the created probability table, agents acting in the home repeatedly perform contact behavior in the home, and the risk of infection is estimated from the amount of virus attached.
Model Settings
Virus transmission rate
・Dry-state virus
・Experimental measurement of influenza virus for each surface material: "contaminated surface → skin (model skin)" and "skin (model skin) → contaminated surface"
・Surface material
・Stainless steel, polypropylene, ceramic, wood (veneer), cardboard, wallpaper, model skin, cotton cloth
・Contact object and material
・207 objects including Call bell (polypropylene), gate (stainless steel), outdoor clothing (cotton cloth)
・Assumed that the percentage of skin contact surface is 100%.
・Line of action
・Categorized by household composition, housing type (apartment, house), floor plan, and panel attributes (e.g., female in her 20s)
・Place of visit (front door, living room, etc.), number of visits, order of visits
・Location and contact objects (keys, doorknobs, light switches, etc.)
・Analysis of visit probability, conditional visit probability, contact order, and number of contacts by location, etc.
Experimental Results
・Visualization of contact infection risks at home after returning home from going out
A synthetic persona created from the acquired data performed movement and contact behaviors based on probability after returning home, reproducing situations in which viruses attached to hands and belongings while going out spread to items in the home, and the contact objects were observed.
・Test results
Virus-attached contacts (maximum risk) within 30 minutes of returning home were found in a wide range of areas, including the entrance area and kitchen.
Experimental Results
・Test results Dry-state virus
The top items in terms of the average amount of virus adherence were the main gate (entrance)_mail and parcel delivery, inside the entrance_disinfectant, outside the entrance_numeric keypad, inside the entrance_bags and shopping bags, inside the entrance_key, kitchen_smartphone, and inside the entrance_doorknob.
Summary
1. The model visualized contact infection risks at home after returning home from going out.
2. The majority of virus adherence is concentrated near the main gate and front door area.
3. In particular, the following items posed a high risk: mail and parcel delivery in the main gate (entrance) area, disinfectant as well as bags and shopping bags inside the entrance, and numeric keypads outside the entrance.
4. It suggests that washing hands before entering the main gate (entrance) or front door (before touching keys, mailboxes, etc.) and disinfecting shopping bags and smartphones, which are frequently touched outside, can reduce risk.
5. We plan to refine the behavior patterns and contact surface ratio in the layout.
