Abstract:
The growing global elderly population and rising incidence of mobility-limiting
injuries have significantly increased the number of bedridden patients requiring
long-term care. Caring for these individuals in hospitals and home settings imposes
substantial physical and emotional burdens on caregivers, while commercially
available smart medical beds remain prohibitively expensive and inaccessible
in resource-limited environments. This study presents a low-cost smart medical
bed that integrates millimeter-wave (mmWave) radar-based non-contact vital
signs monitoring with bilingual (Sinhala and English) voice-activated bed control
to enhance patient autonomy and safety.
The system employs a 60 GHz mmWave radar sensor for continuous, noninvasive
monitoring of heart rate and respiratory rate, eliminating the need for
attached physiological sensors. An ESP32-based voice control unit, connected to
a cloud speech recognition service, enables patients to adjust bed positions via
Sinhala and English voice commands, while a Raspberry Pi manages radar signal
processing, data logging, and communication with a mobile application that
displays real-time and historical vital signs. Experimental validation under controlled
conditions demonstrated 95% accuracy in heart rate and respiratory rate
compared with a medical-grade pulse oximeter at a 1.5 m sensing distance, with
minimal body movement. Voice command recognition achieved 95% accuracy for
English (38/40 successful commands) and 80% accuracy for Sinhala (32/40 successful
commands). The complete system was implemented at a cost below LKR
70,000, offering over 90% cost reduction compared with typical smart medical
beds. The proposed solution, therefore, provides a cost-effective, technically robust
assistive bed platform suitable for home care and rural healthcare settings,
with the potential to reduce caregiver workload and improve continuous monitoring
of bedridden patients.
