What to Leave Behind in Places with Decreasing Populations

In Hiroshima Prefecture, there are 53 areas without doctors, the second highest number in Japan after Hokkaido. Over the past three years, the number of unmanned JR stations has increased by 30. Meanwhile, in Etajima City, a pilot project utilizing AI for welfare support is underway.

At first glance, these three phenomena may seem unrelated. However, when viewed from a broader perspective, a common question emerges: “How can we maintain the functionality of life in places where people have disappeared?”

In response to this question, what is currently happening in Hiroshima Prefecture is not merely the introduction of technology or the replenishment of human resources, but rather the exploration of a “system” that connects the two. Who is this system meant to ease? What can be replaced, and what cannot? We will trace its current state from the tangible experiences on the ground.

—

53 Unmanned Medical Areas: The Issue of the “Last Person” Behind the Numbers

According to the latest survey by the Ministry of Health, Labour and Welfare (2022), there are 601 areas without doctors nationwide. The 53 areas in Hiroshima Prefecture rank second after Hokkaido’s 63 areas. However, the circumstances in Hokkaido and Hiroshima differ. In Hokkaido, the vast area leads to dispersed medical zones, while Hiroshima’s mountainous regions and islands create access gaps.

Many residents living in these unmanned medical areas are elderly. According to prefectural data, it is not uncommon for the aging rate in these areas to exceed 50%. A distance of “40 minutes by car to the nearest clinic” effectively translates to a medical void for elderly individuals who have surrendered their driver’s licenses.

In response to this challenge, the prefecture has been promoting the introduction of online medical consultations and the expansion of mobile clinics. In the fiscal year 2023, trials for online consultations began in five areas within the prefecture. However, there is a problem that is often overlooked. To participate in online consultations, someone is needed to assist with operating the devices. Someone is needed to read the values from blood pressure monitors and pulse oximeters and communicate them to the doctor. In other words, even with technology, if there is no person to carry out the “final step” of delivering it, the system will not function.

A doctor from a clinic in northern Hiroshima, which continues to provide mobile consultations, explains, “Coming just once a month does not reveal the daily lives of patients. It is the local people who know their usual conditions that help improve the accuracy of medical care by noticing when something seems off.” The quality of medical care is not solely supported by the doctor’s skills. It is the community’s eyes that observe daily life and pick up on anomalies—this structure is what barely sustains medical care in unmanned areas.

—

30 New Unmanned Stations: The True Cost of “No Station Staff”

As part of its efforts to improve operational efficiency, JR West has been unmanning stations with low usage. In Hiroshima Prefecture, approximately 30 stations have been unmanned over the past three years. According to JR West’s announcement for the fiscal year 2023, the ratio of unmanned stations in its jurisdiction has already exceeded 70%.

What happens when a station becomes unmanned? There is no one to respond when ticket machines malfunction. There is no one to assist wheelchair users with boarding and alighting. There is no one to speak up if there are suspicious individuals. While this may not be reflected in the numbers, the functions that “having people at the station” provided were not limited to ticket sales.

At an unmanned station on the Kure Line, local residents have voluntarily continued to clean the station building. They plant flowers and post handwritten guides next to the timetable. “If there’s no one at the station, visitors from outside might feel uneasy,” says a woman in her 70s who continues the cleaning. This act is more about “leaving a human presence in the place” than maintaining transportation infrastructure.

On the other hand, the increase in unmanned stations has accelerated discussions on alternative transportation. In Hiroshima Prefecture, demand-responsive shared taxi services are expanding in cities like Miyoshi and Shobara. Users make reservations in advance, and AI calculates the optimal routes for operation. In the case of Miyoshi, the fare per ride is about 300 to 500 yen. The annual operating cost is estimated at around 12 million yen, but compared to the deficit subsidies for route buses, the cost-effectiveness is considered high.

However, there is also a “connecting” issue here. Reservations are made via smartphone or phone call, but some elderly individuals find “making a phone call itself to be a hassle.” A representative from Miyoshi City reveals, “In many cases, community welfare commissioners make reservations on behalf of the elderly. While this was not anticipated in the system, it is how things are functioning in reality.” The movements of people, which are not included in the system’s blueprint, are what enable the system to operate.

—

AI Welfare in Etajima: Can Technology Become a “Listener”?

Etajima City has a population of about 20,000, with an aging rate of approximately 45%. Starting in the fiscal year 2024, a pilot project utilizing AI for monitoring and welfare consultations for the elderly will commence. Through tablet devices, AI will listen to daily health conditions and concerns and connect users to welfare personnel or medical institutions as needed.

What is noteworthy in the design of this project is that the role of AI is limited to “listening and sorting” rather than “judging.” AI will not make diagnoses. It will ask standardized questions like, “How are you feeling today?” and “Do you have any concerns?” and detect changes in the content or patterns of the responses, sending alerts to human staff. The judgment and response will be carried out solely by humans.

A representative from the welfare department in Etajima City explains, “When elderly individuals are truly in need, they often cannot voice their concerns themselves. Having a regular system that ‘asks’ allows us to pick up on small SOS signals. AI is to be a listener.”

Six months after the pilot began, there are about 80 participants. Interestingly, it is not the dialogue with AI itself that builds trust among participants, but rather the experience of “what I talked about with AI led to a visit from the welfare staff the next day.” This sense of having a person behind the technology, rather than trust in the technology itself, is what supports the establishment of the system.

However, challenges remain. About 30% of participants are not accustomed to operating tablets, and local volunteers visit once a week to provide support on how to use them. Here again, the presence of individuals bridging the gap between technology and residents is what distinguishes the success of the system.

—

The Structure Indicated by Three Sites: Designing the “Intermediate Layer”

The community’s eyes supporting mobile medical consultations in unmanned areas. The welfare commissioners effectively facilitating reservations for alternative transportation at unmanned stations. Volunteers assisting with tablet operations for AI welfare.

What is common among these three sites is the structure where an intermediate presence, not explicitly stated in the “blueprints” of the system or technology, ensures the system’s effectiveness. This intermediate layer is neither a professional nor a complete amateur. They are individuals who live in the community, visible at a close distance, taking on small tasks.

This structure possesses both beauty and precariousness. A system that relies on goodwill will collapse if that person is no longer present. The average age of welfare commissioners supporting demand-responsive transportation in Miyoshi City is in the early 70s. Many of the volunteers in Etajima City are also in their 60s or older. The day will come when the “connecting roles” themselves will eventually become the ones needing support.

Therefore, what is being questioned now is not merely “increasing the number of connecting roles,” but rather designing both a system that can operate without connecting roles and a system that can reproduce connecting roles. The former lies in the realms of AI and automation, while the latter pertains to community relationships and education. Neither one alone is sufficient.

—

Future Focus: How to Measure the “Sustainability” of the System

All three examples are still in the experimental stage. They are at a soft phase that cannot be definitively labeled as success or failure. Therefore, the evaluation criteria moving forward will be crucial.

The following three points should be noted:

First, transparency of costs. The annual cost of 12 million yen for demand-responsive transportation, the costs of distributing and operating tablets for AI welfare, and the expenses for establishing communication environments for online consultations—if it is not clear “whose budget this is and how long it will continue,” residents cannot rely on it with peace of mind.

Second, visibility of the intermediate layer. Officially positioning the individuals currently supporting the system within the framework of the institution. If left under the name of volunteers, the burden will concentrate on individuals, leading to burnout.

Third, recording residents’ voices. It is essential to capture not only feedback like “it has become convenient” or “it has been helpful” but also how many voices of “it is difficult to use” or “it is hard to ask for help” can be gathered. The seeds for improving the system lie within dissatisfaction.

—

Introducing technology into places where people have disappeared is not enough to revive them. It is only when someone stands between technology and people, translating the languages of both, that the system becomes a living entity.

The 53 unmanned medical areas, 30 unmanned stations, and small experiments in Etajima—all quietly teach the same lesson. The distance between those who design the system and those who live within it. Ultimately, it is someone’s “small effort” that fills that distance. Naming that effort and providing a place for it—this may be the most necessary design at this moment.