Contactless HMIs with integrated AI help clinics meet hygiene requirements, minimize downtime risks, and ensure data protection. The MediaTek Genio 700 platform, in conjunction with a medical HMI, combines powerful edge AI processing with camera, audio, and graphics capabilities in an industrial-grade design. In clinical applications, it enables secure authentication, voice- and gesture-based operation, and centralized management and maintenance of medical devices—all locally, without cloud dependency.
Introduction: Why clinicians need touchless HMIs with local AI
In clinical environments, hygiene and operational safety are top priorities. Conventional user interfaces with buttons or touchscreens are difficult to clean and provide hiding places for germs, a serious risk for infection control. Contactless human-machine interfaces (HMIs) that use gestures, voice, or facial recognition enable sterile operation without compromising aseptic working practices.
At the same time, patient-oriented systems require maximum reliability and real-time capability. Local AI processing, called edge AI, ensures responsive control even in the event of network failures and protects sensitive health data from external access. This creates a secure, hygienic, and high-performance interaction between humans and machines, tailor-made for the requirements of modern medicine.
MediaTek Genio700 platform delivers high power for AI inferencing
A powerful edge platform is crucial for use in contactless medical interfaces. One possible platform is the System-on-Module SOM-SMARC-Genio700, incorporating the MediaTek Genio 700 high-performance edge AI platform. This eight-core processor features two Arm Cortex-A78 cores running at 2.2 GHz and six Arm Cortex-A55 cores running at 2 GHz, and is manufactured using energy-efficient 6 nm technology. An integrated Neural Processing Unit (NPU) with up to 4 TOPS of performance and a vision processing unit (VPU) enable real-time AI inferencing for gesture and object recognition. The Mali-G57 graphics supports dual displays up to 4K30/60 fps and ensures smooth, detailed user interfaces.
In addition, to capture sensor data, the platform features two image signal processors (ISPs) that can handle cameras up to 32 MP or dual 16 MP setups. Functions such as high dynamic range (HDR) color channels, noise reduction, and autofocus ensure accurate detection under changing lighting conditions. This is complemented by a dedicated audio Digital Signal Processor (DSP) and extensive interfaces for microphones and speakers, ideal for voice control and acoustic feedback.
With versatile I/O options, including USB 3.2, Ethernet with TSN, PCIe, I²C, or SPI, as well as industrial-grade temperature resistance from -40 to +105 °C, the SOM is designed for continuous operation in clinical environments. Operating systems such as Android, Yocto Linux, and Ubuntu offer developers lots of flexibility. In a contactless HMI architecture, the module handles AI processing, controls displays and sensors, and connects securely to local networks, a compact, scalable building block for smart, hygienic medical devices.
The platform enables a secure, responsive, and hygienic HMI solution
An example shows how an AI-supported, contactless workflow based on the Genio 700 supports medical staff in everyday hospital life. The modular, fanless Modular Vision HMI, which incorporates the SOM-SMARC-Genio700 from SECO, integrates a display, plus an external camera and microphone array, positioned close to the patient to make vital data and treatment information accessible. After starting Yocto Linux or Android, the system connects to the hospital network and loads the necessary AI models for speech and image recognition, and provides an intuitive user interface with “Smart Clinical Record” and “Alert Management” scenarios.
As soon as a specialist enters the workplace, the connected camera processes the image data locally and identifies the user via facial recognition and opens a personalized dashboard with the respective access rights. In Smart Clinical Record Mode, patient information such as vital signs, medication plans, or treatment histories is then displayed. Changes, e.g. of the infusion rate, are made via voice commands or gestures, which the Genio 700 processes entirely locally thanks to its audio DSP and NPU, without transferring data to the cloud. For example, if the user says, “Set the infusion rate to 50 milliliters per hour,” the integrated AI converts the command into text, interprets it, and updates the patient file in real time. The adjustment is immediately highlighted on the display.
The extended alert management scenario demonstrates the technical service flow: The HMI monitors devices connected via Ethernet or I²C, such as tomographs or ventilation systems, displays temperature, power, and system status data, and detects anomalies in real time. In the event of malfunctions, technicians are authenticated via notification and receive step-by-step instructions on the display to perform maintenance in accordance with regulations.
After completing care or maintenance tasks, the system synchronizes the updated data with the hospital information systems. All actions are logged for traceability and audits, and sensitive information is transmitted in encrypted form. In standby mode, the vision engine can monitor the room, detect movement, and trigger an alarm if necessary. Features such as automatic identification, personalized dashboards, contextual notifications, and remote maintenance support demonstrate how the Genio platform enables a secure, responsive, and hygienic HMI solution for hospital operations.
Added Value: Device Management and Extensions
Genio-based HMIs can be managed centrally via a cloud-based portal implemented with SECO's Clea software framework. Administrators can roll out new AI models, adjust configurations, or perform software updates on multiple devices simultaneously. Integrated asset management provides real-time insight into the status, runtime, and performance of the systems, a decisive advantage for minimizing downtime and efficiently meeting regulatory requirements.
Thanks to Genio's versatile interfaces, additional sensors such as thermal imaging cameras, RFID readers, or medical scanners can be easily integrated. Edge-to-cloud frameworks allow developers to run customized AI applications, such as anomaly detection or sample monitoring, directly on existing hardware. Optionally available communication modules, including Wi-Fi 6, BT, or 4G/5G, also enable seamless integration into wireless medical facility networks.
Genio-powered medical demo at embedded world North America 2025
The use case described above has been implemented in a practical demo application that was developed specifically for use in clinical environments. The system shows how AI-supported, touchless operating concepts can be seamlessly integrated into existing medical facility workflows using the Genio 700 platform.
In the “Smart Clinical Record” demo, visitors of embedded world North America (EWNA) can call up vital data via gesture or voice control, while the “Alert Management” scenario simulates maintenance and troubleshooting on medical devices. This combined demonstration will be presented exclusively for the first time at EWNA at SECO booth #7011, offering trade visitors the opportunity to experience the complete workflow – from user recognition to real-time AI processing.
Conclusion
Medical facilities can utilize contactless user interfaces that run AI models locally to reduce infection risks, ensure reliability, and protect patient data. The SOM-SMARC-Genio700 offers the necessary performance: an octa-core processor, a 4-TOPS NPU, and industrial robustness form the basis for AI-supported control directly on the device.
In a contactless HMI architecture, the Genio 700 handles AI processing of camera and audio data, controls the display, and communicates securely with hospital systems. With this setup, authentication, voice and gesture control, as well as device monitoring, can be implemented locally and reliably. With central management options, remote maintenance, and flexible expandability, the platform is ideal for scalable, hygienic medical HMI systems.
Contact us for more information on SOM-SMARC-Genio700.