Event Report

Kokoiko Evacuation Drill 2026

Held on Sat, Oct 31, 2026 Event finished

Organizer Kim Laboratory, School of Data Science, Yokohama City University
Co-organizer Kanazawa Ward Self-help Liaison Council (Jijo-Kana)

On Saturday, October 31, 2026, we held the Kokoiko Evacuation Drill 2026 at Kanazawa Public Hall and Deiki Park in Yokohama.

Participants entered their group's situation (stairs, wheelchair use, mobility needs, slopes) into Kokoiko, our evacuation-support web app, and walked to the evacuation site the app recommended. Some groups also experienced a change of destination on the way.

51 people in 29 groups took part, and every group reached its evacuation site. There were no accidents, illnesses or lost participants.

We sincerely thank everyone who took part, and our co-organizer, partners and volunteers.

Participants gathered at Deiki Park before the start
Deiki Park before the start (October 31, 2026)

51

Participants in the drill

29 groups; 57 people registered in advance

100%

Reached their evacuation site

All 29 participating groups checked in

79%

Said the site and route suited them

4 or 5 on a 5-point scale, 33 survey responses

39→52%

Have decided where to evacuate in a tsunami

Before → after the drill, 4 or 5 on a 5-point scale

Overview

Date
Saturday, October 31, 2026, 9:30–14:00 (sunny)
Venue
Kanazawa Public Hall (check-in and opening) and Deiki Park (start, return and booths), Kanazawa-ku, Yokohama
Organizer
Kim Laboratory, School of Data Science, Yokohama City University
Co-organizer
Kanazawa Ward Self-help Liaison Council (Jijo-Kana)
Partners
Kanazawa Ward Office (a pre-event hearing on the ward's evacuation planning); Kanazawa Police Station; evacuation sites: Yakuo-ji Temple, Hakkei Elementary School and AEON Kanazawa-Hakkei; booth partners and a sponsor (listed on the Japanese page)
Staff
17 staff members (faculty, students, lab alumni, Jijo-Kana members and volunteers)
Program
A field trial under the Yokohama City University FY2026 Faculty Community Engagement Support Program
Opening remarks by the chief of Kanazawa Police Station
Opening remarks by the chief of Kanazawa Police Station
Briefing at Kanazawa Public Hall
Briefing at Kanazawa Public Hall
Walking to an evacuation site
Walking to an evacuation site
Check-in at an evacuation site
Check-in at an evacuation site
Check-in at Hakkei Elementary School
Check-in at Hakkei Elementary School
A booth at Deiki Park
A booth at Deiki Park

Evacuation sites and travel times

Assuming a tsunami, each group left Deiki Park at 11:00 and followed the app to one of three sites: Yakuo-ji Temple (high ground), Hakkei Elementary School or AEON Kanazawa-Hakkei. Two minutes after the start, groups able to handle slopes and longer walks were rerouted from AEON to Yakuo-ji Temple as a training scenario.

Site reached Groups Arrival Time from the 11:00 start
AEON Kanazawa-Hakkei Groups matching none of the four input items 6 11:07–11:11 7–11 min (median 10)
Hakkei Elementary School Groups with someone needing mobility consideration 10 11:10–11:17 10–17 min (median 11)
Yakuo-ji Temple (direct) Groups with someone who cannot climb stairs 2 11:20–11:21 21 min
Yakuo-ji Temple (rerouted from AEON 2 min after the start) Groups able to handle slopes and longer walks 14 11:13–11:26 13–26 min (median 21)
Hakkei Elementary School (chose a site other than the recommended one) Same as above 1 11:15 15 min
Map of Deiki Park (start) and the three evacuation sites (labels in Japanese)
The start point and the three sites; labels in Japanese (map: © OpenStreetMap contributors / Protomaps)
  • The table covers the 33 groups with check-in records. 29 groups (51 people) took part in the drill; 4 records were submitted in a different format and were not matched to registered groups.
  • Times run from the 11:00 start to each group's first check-in. Because check-in was done after arrival with staff guidance, actual arrival was 1–2 minutes earlier.
  • The arrival points were for the drill only and differ from the actual places of refuge: the upper floors at Hakkei Elementary School and the 5th-floor rooftop parking at AEON Kanazawa-Hakkei.
  • The destinations applied only to this drill's scenario. A different earthquake would mean different destinations.

Survey results

We received 33 responses (about 65% of the 51 participants), answered on smartphones after returning. Responses are per person and include several from the same group.

Have you decided where to evacuate in a tsunami?

n=33, 5-point scale; numbers are respondents

Before the drill 4–5: 13 (39%)
After the drill 4–5: 17 (52%)
  • 1 Not decided
  • 2
  • 3
  • 4
  • 5 Decided

How participants rated the guidance

n=33, 5-point scale; numbers are respondents

The site and route suited me and my companions 4–5: 26 (79%)
The guidance in the app was easy to follow 4–5: 17 (52%)
  • 1 Disagree
  • 2
  • 3
  • 4
  • 5 Agree

What was hard to follow

19 respondents, multiple answers allowed; top five

  • Which way to go or where to turn 7
  • Where I currently was 5
  • Why the guidance changed 5
  • Looking at the phone while walking felt unsafe 5
  • The connection dropped 3

Features participants would like

32 respondents, multiple answers allowed; top five

  • Whether a facility is open right now 22
  • Where family members are 18
  • Voice guidance 17
  • Works without a connection 14
  • Other hazards (heavy rain, floods) 13

Asked about joining the next drill, 19 people (58%) said they would, 14 said it depends on the content, and none said they would not.

Findings

  • Attribute-based guidance was accepted

    79% said the site and route suited them and their companions, and nobody said they did not. A group with a wheelchair user, accompanied by a caregiver, reached the stair-free high ground (Yakuo-ji Temple) in 21 minutes. Sidewalk width, steps and slope gradient were raised as issues.

  • How the guidance is shown needs the most work

    Only 52% found the guidance easy to follow. The reasons centred on three points: the direction to walk, the current position, and why the guidance changed.

  • Not everyone noticed the change of destination

    Groups rerouted on the way arrived at about the same time as those who went directly. However, 4 of the 15 people whose destination changed did not notice the change by themselves. Notifications and a clear reason are needed.

  • The closing session could not be held

    Arrival times ranged from 11:07 to 11:26 depending on the site, and the return times spread out even more, so the wrap-up for all participants was not held. The next event will be planned around staggered returns.

Next steps

  • Improve the guidance screen: a map oriented to the walking direction, reliable position updates, reasons for recommendations and changes, and sound or vibration alerts
  • Survey slopes, stairs and sidewalk widths on foot, and compile data on site conditions, elevation and gradient
  • Work with the ward and facilities on a way to reflect whether a site can actually be entered at that moment
  • Aim to guide people along pre-registered sites and routes even without a connection
  • A drill is planned in Kanazawa Ward in October 2027, with expansion to the Namiki and Sachiura areas under consideration

Contact

For questions about this event or the report, please email us at:

ahyikyokohama-cu.ac.jp

* Displayed in a non-copyable form to prevent spam.

About the Kokoiko project →

Would your community group, school or workplace like to hold its own drill with Kokoiko? See the call for groups (Japanese) →