The construction of community elderly care facilities (CECF) is pivotal for promoting healthy aging and “aging in place” for older people. This study focuses on the low utilization rates of community elderly care facilities in the Dongcheng and Xicheng Districts, core areas of Beijing. The explainable machine learning method is used to analyze data across three dimensions: the elderly’s individual attributes, characteristics of the community elderly care station (CECS), and features of the built environment around CECS and subdistrict, to identify the important factors that influence the usage frequency of overall CECS and its different functional spaces, and also the correlation between factors and usage frequency of CECS. It shows that the most important factors are the features of CSCF, including the degree of space acceptance and satisfaction with services provided, which influence the usage frequency of nine functional spaces (R² ≥ 0.68) and overall (R² = 0.56). In addition, older people’s individual factors, such as age and physical condition, significantly influence the usage of specific spaces such as rehabilitation therapy rooms and assistive bathing rooms. The influence of built environment characteristics is relatively low, with factors such as the density of bus stations and housing prices within the subdistrict and the mean distance from CECF to the nearest subway stations being more important. These findings provide a reference for the construction of indoor environments, management of service quality, and optimal site selection for future community elderly care facilities.

China, like many other countries around the world, faces a rapidly aging population. “Aging in place” is the most popular choice for older people in China, with more than 90% of urban older people living in the same residential areas they did in middle age. Nighttime outdoor activity (NOA) is popular among the elderly and has a positive impact on both their physical and mental health. However, the conditions of nighttime activity places in old gated communities often do not meet the activity needs of the elderly, and there are problems such as low lighting and poor accessibility. This study focuses on typical old gated communities in a large city and analyzes data on three dimensions of independent variables: the lighting of activity places, the built environment of old gated communities, and the attributes of the elderly residents themselves. Taking “satisfaction with the environment for NOAs”, “the biggest environmental problem with NOAs”, and “locations that need improved lighting” as dependent variables, we used machine learning to calculate the contributions of various influencing factors on the dependent variables. It was found that the most important factors that influence satisfaction with the NOA environments are lighting levels, with the built environment or elderly residents’ attributes being less important. The most important factor that influences older people’s judgment of “locations that need improved lighting” is the perception of safety related to lighting in NOAs, followed by the perception of uniform lighting and green areas in their gated community. This study provides a reference for the renovation of age-friendly community environments as well as the overall enhancement of the nighttime activity environments in old gated communities.

China’s population is aging rapidly, with a large proportion of elderly individuals “aging in place”. In central areas of large cities, the amount of community and home-based elderly care services provided by the government and for-profit organizations are insufficient to meet the demands of these “aging in place” elderly. Taking the core area of Beijing as the spatial scope, this empirical study collects the demand on services of the main types of elderly residents in community and home-based dwelling through questionnaires (n = 242) and employs a mixed-methods approach for analysis. Descriptive statistics and exploratory factor analysis are used to determine the categories and levels of those demands, and machine learning methods (random forest regression model) are used to calculate the importance of various influencing factors (features of the elderly and subdistricts’ built environment) on them. It is shown that elderly residents have a higher demand for psychological and physical condition maintenance services (mean = 3.40), and a lower demand for reconciliation and rights defense services (mean = 3.08). The results also show that the built environment factors are very important for the elderly on choosing demands, especially mean distance of CECSs (community elderly care stations) to downtown landmarks and main roads in subdistricts, and characteristics of CECS. The elderly’s own features also have a relatively important impact, especially their living arrangements, caregivers, and occupations before retirement. This study applies machine learning techniques to sociological survey analysis, helping to understand the intensity of elderly people’s demand for various community and home-based elderly care services. It provides a reference for the allocation of such service resources.

Several methods have been applied to measure healthcare accessibility, ie, the Euclidean distance, the network distance, and the transport time based on speed limits. However, these methods generally produce less accurate estimates than actual measurements. This research proposed a method to estimate historical healthcare accessibility more accurately by using taxi Global Positioning System (GPS) traces. The proposed method’s advantages were evaluated vis a case study using acute myocardial infarction (AMI) cases in Beijing in 2008. Comparative analyses of the new measure and three conventionally used measures suggested that the median estimated transport time to the closest hospital with percutaneous coronary intervention (PCI) capability for AMI patients was 5.72 minutes by the taxi GPS trace-based measure, 2.42 minutes by the network distance-based measure, 2.28 minutes by the speed limit-based measure, 1.73 minutes by the Euclidean distance-based measure; and the estimated proportion of patients who lived within 5 minutes of a PCI-capable hospital was 38.17%, 89.20%, 92.52%, 95.05%, respectively. The three conventionally used measures underestimated the travel time cost and overestimated the percentage of patients with timely access to healthcare facilities. In addition, the new measure more accurately identifies the areas with low or high access to healthcare facilities. The taxi GPS trace-based accessibility measure provides a promising start for more accurately estimating accessibility to healthcare facilities, increasing the use of medical records in studying the effects of historical healthcare accessibility on health outcomes, and evaluating how accessibility to healthcare changes over time.