HOMEWORKS / Park AI model日本語
ANALYZE / CONNECT ONGOING

Park AI model
Verifying the Impact of Green Space, in Numbers

Turning intuition and experience
into something that can be handed down.
CANOPY — greenery ratio, dwell time, flow / AI camera SOIL — moisture, temperature / sensor Growth anomaly detected
PROJECT
Joint venture with Logian and WaveCore (three Kyushu University-founded startups)
TRACK RECORD
Adoption confirmed for a section within the GREEN×EXPO 2027 site
YEAR
2026 — ongoing
SCOPE
Foot-traffic and dwell-time analysis via AI camera / Environmental sensors / Impact verification through causal inference / Integrated dashboard
01 Background

Green space
has never been measured.

Maintaining planted greenery has long depended on the rules of thumb held by veteran gardeners. As their numbers shrink and management budgets shrink along with them, operations built on the current structure cannot continue as they are. At the same time, we have had no way to put a number on how the green space we build actually benefits visitors and the surrounding community — the kind of evidence needed for budget requests, resident briefings and investment decisions.

On top of that, foot-traffic analysis, image analysis, sensors and signage are each ordered from different vendors, so costs stack up while the data stays scattered across separate systems and is never put to use. These look like unrelated problems, but they all trace back to a single cause: nothing is actually being measured. What's needed now isn't more staff — it's turning rule-of-thumb experience into data the next person can pick up and carry forward.

02 Proposal

Measure the site,
and act only when it matters.

We provide AI cameras, environmental sensors, signage and an analytics platform together through a single point of contact. Bringing what used to be disconnected pieces onto one platform removes duplicated cost and lets management decisions weigh growing conditions against how the space is actually used.

The reason three companies work together: it lets us run
one unbroken line from data capture to decisions to on-site delivery.

Other companies can install sensors. Others count foot traffic, and others run statistical analysis. What three Kyushu University-founded companies bring by working together is the ability to handle everything from data capture through to convincing decisions and on-site DX without a break in the chain. URBANIX analyses the foot-traffic and dwell-time data from the AI cameras and serves as lead company and point of contact; Logian handles impact verification through causal inference and KPI design; WaveCore implements the outdoor sensors and cameras.

03 Who It's For

For park managers,
and for developers investing in green space.

The same measurement platform answers two quite different needs.

04 Data Handling

What the cameras capture:
only how many people, and how they move.

Installing cameras at all means we need to be upfront about designing the system so it cannot identify individuals. We do not use facial recognition to identify anyone; each person is treated only as a temporary ID used solely to tally movement. Footage is, in principle, not stored in any form that could identify an individual, and is deleted automatically once a set retention period ends.

We post signage stating that filming and measurement are taking place, together with the purpose and a contact point, and restrict data access to authorised staff only. Data collected is never provided or sold to third parties. Operating rules are agreed and put in place before launch, in line with Japan's Act on the Protection of Personal Information and the client's own regulations and ordinances.

05 Where We Are Now

Not stopping at an exhibit —
built to carry through to long-term operation.

RELATED WORKS