Hypercustomization and Seamless Integration with Legacy Infrastructure Critical for Seeed Studio to Succeed in the Enterprise Edge AI Market

Shenzhen-based Seeed Studio announced in March 2022 that it would start to offer end-to-end edge Artificial Intelligence (AI) solutions for enterprises. Seeed Studio is well known in the developer community as the hardware module manufacturer of the Internet of Things (IoT) and the maker industry. In addition, the company readily supports hardware projects from hardware design to embedded software design, packaging design, manufacturing, and certification, making it easy for developers to prototype, integrate, and deploy their edge AI projects quickly.

In March 2022, Seeed Studio announced its first batch of products targeted at edge AI applications. reComputer is an edge AI device that features NVIDIA Jetson Nano for low-power Machine Learning (ML) inference. In addition, the box comes with various Input/Output (I/O) options to support expansion and connectivity. Meanwhile, reServer is an on-premises computing server with a more powerful NVIDIA AGX Orin, rapid network access, and hybrid connectivity. Both edge AI hardware are pre-installed with the NVIDIA Jetpack featuring the NVIDIA Jetson software stack to facilitate edge AI deployment.

At the moment, Seeed Studio has already developed several custom edge AI solutions for various software partners and system integrators. For example, the company helps Zenus AI develop a smart camera that can analyze customer impressions, demographics, and sentiments in smart retail. The camera is designed to perform on-device inference using NVIDIA’s Jetson Xavier NX, ensuring data privacy and data sovereignty compliance. In industrial settings, reComputer can be deployed to detect workers’ Personal Protection Equipment (PPE), such as safety helmets and mask detection. In traffic management, the company is partnering with alwaysAI working on an AI application at the edge for pedestrian, vehicle, cyclist, and license plate detection. Finally, in robotics, Seeed Studio is partnering with Cogniteam to create an end-to-end solution for robotics design and operation.

At the same time, the company has also developed a full stack solution for environmental monitoring. The company offers sensors, gateways, servers, databases, a web portal and dashboard, analytical tools for smart agriculture and farming, weather and environmental monitoring, safety and security monitoring, and asset management. In addition, the company is also working on its open-source zero-code AI vision tool.

According to ABI Research’s IoT forecast, the global IoT connections are expected to grow from 8.3 billion in 2021 to more than 23 billion in 2026. At the same time, the global edge AI chipset shipments are expected to exceed 5 billion in 2026, with strong growth coming from TinyML and edge AI gateways and servers. There are plenty of opportunities in the edge AI space if Seeed Studio positions itself correctly. While many companies have started to have developed strong institutional knowledge of their AI use cases, edge AI operation remains a considerable challenge. In addition, supporting a global base of customers from design to commercialization and monitoring around the clock will be a huge strain on the 300-employee-strong Seeed Studio.

Having the right partnership is an excellent place to start. Seeed Studio has partnered with several innovative edge AI solution providers and system integrators to lower the barrier to entry into edge AI. Solution providers like Edge Impulse, Deci, alwaysAI, and Cogniteam provide innovative low-to-zero code tools for data collection, model training, optimization, validation, and monitoring. At the same time, system integrators like Tryolabs and Zenus AI accelerate edge AI model deployment for specific applications, such as asset tracking, facial analysis, facial recognition, and predictive maintenance, while providing support on monitoring, upgrading, and maintenance.

It will be interesting to see how Seeed Studio will further differentiate itself from its competitors, many of which are established software vendors or cloud solution providers. Also, transitioning from a developer-centric business model to an enterprise-centric one will take significant effort. Enterprises are looking for companies with industrial experience and a strong track record in serving the enterprise sector. As a company that specializes in serving the long-tailed maker industry, Seeed Studio must be able to introduce the same level of hypercustomization to hardware and software for large enterprises. The company also needs to ensure its solutions can be integrated seamlessly with legacy infrastructure, which is critical to existing enterprise workflow.