According to an October 2023 survey by ManpowerGroup, as many as 75% of employers reported staff shortages, showing that labor shortages are a widespread industry challenge. In addition, amid the global trend toward net-zero carbon emissions, many major companies have begun to focus on carbon footprint reporting across their supply chains. Pain points such as labor shortages and energy management needs have driven the industry to actively seek smart solutions. Now that artificial intelligence (AI) has become a hot topic, hardware devices that combine AI models and algorithms can help businesses overcome these challenges.
DSI CEO Tsai Yu-Che explains that the labor shortage, combined with growing data volumes and rising technical demands, is the underlying reason automation has become an essential requirement. Take factories as an example: technicians who have worked for decades are struggling to find successors, while production density and technical capability keep increasing, making it necessary to adopt automation solutions to support operations. In addition, as international bodies and governments place growing emphasis on carbon emissions, generating carbon inventory reports and building sound energy management systems has become an urgent priority. Smart devices with data collection and processing capabilities can help businesses obtain power usage data and optimize energy management through automated functions such as time-of-use electricity pricing.
These two essential needs reveal a strong industry demand for AI-powered smart applications. With AI models developing rapidly, finding the right AI model and ensuring it runs smoothly on hardware devices is key to building automation solutions. However, most businesses today either adopt existing AI models from major providers such as Google, or invest heavily in training custom models to meet their specific needs. This means AI applications are either constrained by the limitations of off-the-shelf models and fail to address real pain points, or the high cost of training entirely new models discourages adoption.
Tsai believes that simplifying the AI model training process and lowering adoption costs are essential to driving widespread AI deployment and delivering truly impactful change. With this in mind, DSI leverages its unique strength as a "software company with hardware development capabilities" to address these two essential needs, launching an AI edge computing recognition camera that helps businesses improve operational efficiency, as well as a power and energy inventory gateway product that measures, collects, and records electricity usage data in real time. Through its combined software and hardware expertise, DSI helps businesses easily apply AI to solve operational challenges.
DSI CTO Kuan Yu-Kai (Figure 2) explains that, taking AI recognition applications as an example, many AI recognition models can only identify a single type of object (such as license plate recognition), whereas DSI's AI edge computing recognition camera comes with multiple built-in AI recognition models, allowing users to flexibly choose the ones that truly meet their needs. In addition, DSI also offers custom AI model development services—once a model is trained, it is deployed on DSI's existing hardware, enabling the rapid development of AI products with the required functionality.
AI applications span a wide range of scenarios. Tsai gives an example: the company has encountered retail businesses wanting to adopt AI solutions to gain deeper insights from customer behavior. Beyond process automation and energy management, as the cost of training AI models continues to fall, AI applications are expected to rapidly expand into a wide variety of use cases, sparking a genuine smart revolution.
News Source: New Communications Components Magazine
