HomeWorldUSALatin AmericaEuropeAsiaAfricaTV ShowsShowbizTravelLifestyleOpinionSciencePoliticsHealthSportsTechEntertainmentBusiness
Business July 24, 2026

AI-powered smart cane by Agora enhances elderly care

AI-powered smart cane by Agora enhances elderly care

A new partnership between Agora and Lgenie has introduced an AI‑enabled smart cane that lets older adults ask questions and receive assistance through voice while walking. The device uses Agora’s low‑latency conversational AI to support outdoor and everyday use without requiring a smartphone or other screen.

The collaboration reflects a broader trend of integrating conversational AI into physical products across healthcare, smart homes, education, robotics, and mobility, moving beyond chatbots and software applications.

Xiao Dong Feng, head of Agora’s Physical AI Product line, said the smart cane is among the earliest examples of AI moving from screens into the physical world. He noted that the technology adapts to the user instead of forcing users to adapt to it.

Designed for older adults, the cane allows natural voice conversations while walking, eliminating the need to stop and operate a device. Users can ask questions and seek assistance in real time, making the experience seamless.

According to Feng, the main challenge lies not in building an intelligent model but in ensuring the device remains dependable under real‑world conditions. He emphasized that older adults judge a product by its responsiveness at busy intersections or in loud traffic, rather than by the sophistication of its algorithms.

Lgenie incorporated Agora’s Conversational AI Engine, which combines speech recognition, speech activity detection, and full‑duplex voice streaming. This integration supports low‑latency conversations, real‑time interruption handling, and stable audio performance even when network conditions fluctuate.

The deployment achieved an average response latency of about 400 milliseconds, improving conversational stability and interruption handling during user interactions. This performance level enables smoother, more natural exchanges.

Feng highlighted that making conversations feel natural is essential, as users should be able to speak as they normally do without adjusting their behavior. He added that good AI answers questions, but great AI reduces hesitation, allowing users to continue speaking without doubt.

Over time, users who initially spoke cautiously became comfortable interrupting the device, asking follow‑up questions, and changing topics naturally. Trust develops when users stop thinking about the technology and simply engage with it.

AI‑powered assistive devices are expected to play a growing role in aging societies, such as the Philippines, where the proportion of people aged 60 and above is projected to exceed 10% by 2030. Smartphone adoption is expected to reach 92% by that year, with more than half of mobile connections running on 5G networks.

Feng stressed that AI should complement rather than replace caregivers, extending assistance beyond the moments when family members are physically present. He noted that family members provide empathy, judgment, and emotional support that technology cannot replace.

Looking ahead, the next phase of AI development will focus less on increasing model intelligence and more on ensuring AI is dependable enough to integrate naturally into daily life. The best AI, he said, will quietly provide support at moments when it is genuinely useful.

Through its partnership with Lgenie, Agora aims to expand the use of conversational AI across healthcare, education, smart home technologies, mobility, and other connected devices that require reliable real‑time interaction.

Share this article

UMVA MAG

UMVA Mag is your trusted source for breaking news, in-depth analysis, and compelling stories from around the world. Covering politics, business, technology, entertainment, sports, health, science, and more — we deliver journalism that matters.

Independent, Accurate, Unbiased
24/7 Breaking News Coverage
Trusted by Millions Worldwide