As AI moves from screens into physical devices, elder care will become one of the clearest tests of whether technology can feel immediate, reliable, and trusted in daily life.
For much of the past decade, the promise of the Internet of Things was built around connection. Devices became connected to the cloud, mobile apps, enterprise dashboards, and each other. That was an important foundation, but it is no longer enough.
A connected device that only collects data or responds to a basic command now feels increasingly limited. In elder care, this limitation is becoming more visible. Older adults, caregivers, and families increasingly need technology that can understand context, respond conversationally, adapt to changing environments, and provide help at the exact moment it is needed.
This is both a technology shift and a care delivery shift. The next phase of elder care will be defined by how intelligently and immediately devices can interact, rather than simply by the number of connected devices.
The urgency is growing. According to the World Health Organization, by 2030, one in six people globally will be aged 60 or older. By 2050, the number of people aged 60 and older is expected to double to 2.1 billion.
As aging populations place more pressure on families, caregivers, healthcare providers, and social systems, elder care technology must move beyond passive monitoring and become more assistive, conversational, and responsive.
From connected devices to AI-native care experiences
Elder care technology is moving from passive connected hardware toward intelligent, real-time, and responsive care ecosystems. The device is no longer simply an endpoint. It is becoming an AI-native interface for continuous interaction among older adults, caregivers, AI systems, and the physical environment.
The next frontier involves building AI-native devices from the ground up, with sensing, connectivity, speech, reasoning, and response designed as one integrated experience.
In elder care, this distinction has practical consequences. A device that sends an alert may help caregivers know that something has happened. An AI-native care device should also help users and caregivers understand what is happening, determine what action may be needed, and respond in real time.
The business implications are significant. According to McKinsey’s 2025 State of AI survey, 88 percent of respondents said their organizations regularly use AI in at least one business function, while 62 percent said their organizations were at least experimenting with AI agents.
McKinsey also noted that most organizations remain at the experimentation or pilot stage when scaling AI across the enterprise. This gap between adoption and operational impact is particularly relevant to elder care, where AI must work reliably outside controlled digital environments and support real human needs.
The challenge is no longer simply whether companies can build AI into products. Many can. The harder question is whether they can make AI work reliably under real-world conditions, including weak networks, noisy environments, fragmented data, varied user behavior, and situations where delayed or inaccurate responses can quickly erode trust.
In elder care, trust is central to whether older adults and caregivers will rely on the technology in daily life.
The physical AI stack behind real-time elder care
Real-time engagement is becoming a core layer of AI-native elder care devices.
In many AI applications, latency is treated as a technical performance metric. In elder care, it can also become a care, safety, and trust metric. A delayed answer from a chatbot may be inconvenient. A delayed response from an assistive device, remote care tool, emergency support interface, or healthcare device may affect whether the user trusts the product.
For AI to work effectively in elder care, it needs a physical AI stack that connects the device, microphones, speakers, sensors, connectivity, real-time audio processing, automatic speech recognition, large language models, text-to-speech systems, interruption handling, and safety workflows.
This is the layer that allows intelligence to move from a cloud-based model into a physical product that people can speak to, carry, wear, or depend on during their daily routines.
Agora’s role in this stack is to provide the real-time interaction layer that helps AI-native devices feel natural and responsive in the physical world.
Through technologies such as its Conversational AI Engine and Convo AI Device Kit, developers can connect AI models, speech recognition, text-to-speech, audio processing, and device-side capabilities so physical products can support more natural voice-driven interactions.
For elder care, this is critical because the experience cannot feel like a delayed command-and-response system. It needs to feel like support that is available in the moment.
Conversational AI makes this even more important. Voice is one of the most natural ways for people to interact with technology, particularly for older users who may not find app-based interfaces intuitive or accessible.
Voice interaction is also demanding. Users interrupt. They speak with different accents. They pause, change their minds, move between environments, and expect the system to keep up. Conversational AI in elder care must be able to listen, process, respond, and recover with the fluidity of a natural conversation.
The industry is already moving in this direction. Gartner’s 2025 Magic Quadrant for Conversational AI Platforms notes that the conversational AI market is expanding beyond traditional service automation into AI agents and multimodal interactions.
Future interfaces will combine speech, sensing, video, location, device data, and contextual signals to create more adaptive experiences. In elder care, devices may become increasingly aware of a user’s environment, routines, and moments of need instead of waiting for a manual command.
Where elder care starts to feel more human
Early examples of this shift are already emerging.
LGenie, for instance, has applied conversational AI to AI-native assistive devices for elder care, where real-time voice interaction, interruption handling, and reliability in outdoor environments are central to user trust.
With Agora powering its conversational AI infrastructure, LGenie transformed its smart cane into a more responsive AI-powered assistive device. The system achieved an average response latency of 400 milliseconds and improved interruption handling in real-world environments.
The significance of this example lies in what the device represents: elder care technology moving from passive hardware toward AI-native support that can listen, respond, and adapt in the user’s everyday environment.
The broader lesson extends beyond one deployment. AI-native devices become more useful when they can support people within the context of daily life, rather than operating only within a controlled app experience.
This movement is also supported by broader infrastructure trends. IoT Analytics estimates that the number of connected IoT devices will grow from 21.1 billion in 2025 to 39 billion by 2030.
More devices alone will not automatically create better outcomes. Without real-time intelligence, connected devices can produce more data, alerts, and operational complexity.
In elder care, this complexity can increase the burden on caregivers, families, and care providers. Greater value emerges when companies turn connected endpoints into AI-native care touchpoints that are responsive, contextual, and easy to trust.
What businesses need to get right
Companies should begin by designing around care moments instead of device features. The central question should be what decision, action, or support the user needs at a particular moment.
This is especially important in elder care and healthcare, where circumstances can change quickly and users may need assistance during emotional, urgent, or unpredictable situations.
Latency should also be treated as part of the care experience strategy. Speed affects whether an interaction feels natural, whether the user remains engaged, and whether the system can be trusted over repeated use.
In conversational AI, even small delays can make an interaction feel mechanical or unreliable. For older users, this can influence whether they feel supported, confused, or ignored.
AI, device, and care-experience teams also need to work from the same architecture. AI is too often added after the device strategy has already been established. That approach may support simple automation, but intelligent and adaptive systems require deeper integration.
AI-native elder care devices depend on close coordination among hardware, connectivity, edge processing, cloud infrastructure, AI models, data pipelines, privacy safeguards, and user experience design.
Companies should also consider the complete physical AI stack rather than evaluating individual components in isolation. A strong model or capable device cannot deliver the full experience on its own. The model, device, network, audio processing, speech layer, and user experience must work together reliably under real-world conditions.
Trust must be incorporated into the product from the beginning. Deloitte’s 2025 predictions on autonomous generative AI agents suggest that agentic AI is moving beyond pilots in some markets and applications.
As these systems become more capable, companies will need stronger governance covering data privacy, transparency, escalation, human oversight, and failure handling. These safeguards become particularly important when AI systems interact with vulnerable users or operate in physical environments.
Elder care demonstrates why responsible AI must be incorporated into the product experience itself rather than treated solely as a compliance requirement.
Companies should also prepare their operations for continuous learning. The development of an AI-native elder care device does not end at launch. These products improve through real-world usage, feedback loops, model refinement, and operational monitoring.
Product, engineering, customer support, compliance, and business teams will need to work together long after deployment. Companies must also listen closely to caregivers, older adults, and healthcare partners as the technology evolves.
The next generation of elder care technology will be judged by whether it solves real problems with reliability, immediacy, and empathy.
Companies that understand this opportunity can move beyond connected devices and develop AI-native care ecosystems that people can genuinely depend on.
In my view, the future of elder care involves more than adding AI to hardware. It requires a physical AI stack that allows intelligence to become present, responsive, and trusted during the moments when older adults and caregivers need it most.
Xiao Dong Feng is Head of Physical AI Product at Agora.
TNGlobal INSIDER publishes contributions relevant to entrepreneurship and innovation. You may submit your own original or published contributions subject to editorial discretion.
Featured image: Camila Mofsovich on Unsplash