January 28, 2026
Article
With the rapid rise of smart surveillance, autonomous drones, and AI-powered industrial vision systems, the need for intelligent processing at the edge is more pressing than ever. These emerging applications demand real-time image processing, low-latency inference, and robust data privacy, often in environments constrained by power, space, or connectivity. Traditional cloud-based solutions are often limited by latency, bandwidth dependency, and privacy concerns, making them inadequate for time-sensitive, field-deployed systems. This is where edge AI platforms like iWave’s iG-G70M OSM System on Module, based on the MediaTek Genio 1200 SoC, play a crucial role.
Edge-based video processing is becoming a cornerstone of next-generation applications, from aerial drones conducting perimeter surveillance to smart traffic systems monitoring congestion patterns in real time. These systems must handle multiple camera feeds, perform on-device analytics such as object tracking and face recognition, and do so with ultra-low latency to ensure responsiveness. Furthermore, they must operate efficiently in thermally and power-constrained environments, especially in mobile or outdoor deployments. The iG-G70M System on Module empowers these capabilities by offering powerful AI acceleration, camera interface flexibility, and efficient processing, all in a compact form factor. According to a recent 2025 study, Edge AI systems can outperform cloud-based approaches in both latency and bandwidth, reducing response times by up to 80% and minimizing data transfer by processing insights locally.
At the heart of the iG-G70M SoM lies the MediaTek Genio 1200, a high-performance AIoT processor manufactured on a 6nm node. The SoC integrates an octa-core CPU configuration, featuring Quad Arm Cortex-A78 cores paired with Quad Arm Cortex-A55 cores, providing a balance of performance and efficiency that is ideal for complex edge workloads. Its GPU, the Mali-G57 MC5, supports advanced graphics rendering and 4K display output, making it well-suited for camera-intensive and media-rich applications. For AI tasks, the Genio 1200 incorporates a powerful NPU capable of delivering up to 3.9 TOPS and DSP into a single chip. This is further complemented by a dedicated Vision Processor (VP6) and MediaTek’s Deep Learning Accelerators (MDLA), which enable smooth, real-time neural network inferencing. Additionally, the Image Signal Processors (ISPs) support high-resolution camera inputs simultaneously, and the integrated 4K60 video codec provides seamless encoding (HEVC) and decoding (HEVC, AVC, and VP9) formats. All of this is wrapped in a highly efficient silicon architecture, making the Genio 1200 ideal for always-on vision-based edge devices.
The iG-G70M leverages the capabilities of the Genio 1200 and encapsulates them in a compact, production-ready OSM Size-L module in a 45mm x 45mm form factor. The module features multiple MIPI CSI interfaces, enabling simultaneous dual-camera input for applications such as stereo vision or combined RGB and thermal imaging. Its wide operating temperature ranges from -40°C to +85°C allows it to function reliably in harsh outdoor and industrial environments. With high-speed interfaces including USB 3.0, PCIe Gen3, and Gigabit Ethernet, the module supports high-throughput data transfer and real-time connectivity. The compact OSM design not only saves board space but also supports future scalability through standardized edge connector integration.
Drones operate in highly constrained environments where size, weight, and power efficiency are critical. The OSM Size-L (45mm x 45mm) form factor of iWave’s iG-G70M makes it an ideal fit for compact drone chassis, without compromising performance. With support for industrial temperature ranges (−40°C to +85°C), the module can withstand harsh outdoor conditions. Its standardized and modular design also enables developers to swap SoMs without redesigning the entire system greatly accelerating prototyping and scaling. For drone-specific applications, the SoM’s lightweight, energy-efficient design supports extended flight endurance, while its high-speed interfaces (USB 3.0, PCIe Gen3, Gigabit Ethernet) ensure real-time telemetry and payload data transmission. Additionally, dual MIPI-CSI camera support enables stereo vision or multi-sensor fusion, a key requirement for navigation and situational awareness.
AI & Vision Capabilities
NPU for Real-Time Inference: The Genio 1200’s NPU enables tasks such as object detection, tracking, and classification directly onboard the drone.
VP6 Vision Processor + MDLA: Accelerates neural network workloads at low latency, ensuring real-time performance.
Dual ISPs: Support for simultaneous high-resolution camera inputs, ideal for multi-sensor drones that rely on stereo vision or sensor fusion.
Efficiency & Performance
Connectivity & Codec Support
A compelling use case for the G70M module is its integration into an AI-enabled surveillance drone. Such a drone must capture and process video in real time, detect people or vehicles, and generate alerts without relying on cloud infrastructure.
Using the dual MIPI-CSI interfaces and ISP support, the G70M can simultaneously handle RGB and thermal cameras. AI inference tasks, such as object detection using YOLOv5 or MobileNet models, are offloaded to the dedicated NPU. This localized processing not only eliminates the latency and dependency associated with cloud transmission but also reduces bandwidth consumption by transmitting only metadata or relevant alerts rather than raw video feeds. The module’s energy-efficient design ensures it can be deployed on battery-powered drones for extended durations, and its USB gadget modes (UVC for video and ECM for Ethernet over USB) enable easy interfacing with companion devices and ground stations.
iWave plays a strategic role in enabling drone manufacturers to adopt edge AI computing at scale. By offering production-ready SoMs like the iG-G70M, iWave provides the hardware foundation, software ecosystem, and long-term support needed for industrial-grade drone deployment. With expertise in low-power design, camera integration, and rugged system development, iWave helps customers build drones that are lighter, smarter, and more efficient. Additionally, as an ODM partner, iWave supports customers from prototyping through certification and production, ensuring faster time-to-market in a rapidly evolving drone landscape.
Software ecosystem and development support
iWave complements the hardware with a production-ready software stack:
Yocto-based Linux BSP: Provides flexibility and ease of customization for developers.
GStreamer: This pre-integrated framework supports video capture, processing, encoding, and streaming pipelines.
AI SDK: Model conversion, deployment tools, and sample inference apps
USB Gadget Support: CDC-ECM for Ethernet over USB, UVC for video streaming
Camera Tuning Tools: Support for DCC/ISP calibration, V4L2 interfaces
With this full-stack offering, iWave ensures that customers have everything they need – from silicon to software – to build cutting-edge edge AI vision systems.
The iG-G70M, powered by the MediaTek Genio 1200 SoC, delivers a compelling combination of performance, flexibility, and efficiency for camera-centric edge applications. Whether deployed in surveillance drones, smart kiosks, or field robots, the module’s integrated AI processing, advanced camera support, and multimedia capabilities make it an ideal platform for next-generation embedded vision systems. Its industrial-grade durability and compact form factor offer deployment versatility, while iWave’s development ecosystem ensures that customers can bring products to market faster. As real-time, privacy-preserving intelligence becomes a cornerstone of edge deployments, the iG-G70M stands ready to lead the way in enabling smarter, faster, and more secure vision solutions.
For more information or inquiries, reach out to mktg@iwave-global.com
We appreciate you contacting iWave.
Our representative will get in touch with you soon!