Automotive Diagnostic Connectors And Cables
We make OBD connector with terminal by ourselves, soldering type and crimping type are both available. Such as 16pin obd connector. OBD1, OB2, J1939, J1708, J1962, etc. Also molded by different type, straight type or right-angle type. The OBD connector cables used for Audi, Honda, Toyota, BWM, etc. We have wide range of materials source , also we can support customers to make a customized one to replace the original ones.
Automotive Diagnostic Connectors And Cables,Obd Connectors,Reliable OBD Connector,Black OBD Connector,OBD Diagnostic Cable,OBD2 Connectors ETOP WIREHARNESS LIMITED , https://www.etopwireharness.com
AI is often considered the final piece of the IoT architecture. In the future, both hosts and terminals will have varying levels of computing power. Taiwanese manufacturers can leverage this trend to develop specialized applications for IoT terminals and networked devices, shifting away from the traditional low-cost OEM model.
The Internet of Things (IoT) has long been seen as the fourth revolution in the IT industry, following PCs, the Internet, and smartphones. The vast business opportunities brought by the internet have led research institutions to make bold predictions. For instance, BI Intelligence estimated that global net-device shipments would surpass smartphones by 2017, while Harbor Research predicted 10 billion connected devices by 2020, with business opportunities exceeding $1 trillion. By late 2016, AI had become a hot topic in the tech industry, drawing significant attention. Once AI captured the industry's interest, it quickly merged with IoT to form a new term: AIoT. Most market participants believe that AI completes the last missing piece of the IoT puzzle, and their combination will accelerate the intelligence of IT systems, unlocking even greater business potential.
From an overall system perspective, IoT is typically divided into three layers: sensing, communication, and application. Since most mainstream AI algorithms are based on deep learning, they improve through continuous error correction, gradually approaching perfection. However, this process requires significant computational power, making it more suitable for high-level application platforms within the IoT ecosystem. Recently, the concept of edge computing has gained traction, offering a promising opportunity for Taiwan in the AIoT space.
Edge computing presents new business opportunities for Taiwan.
Currently, the IoT relies heavily on a centralized computing architecture, where data from the first layer is uploaded to a cloud platform for storage and analysis. While centralized and decentralized models each have their strengths and weaknesses, their applications differ. Centralized computing faces challenges such as real-time performance, processing load, and transmission costs. For example, in manufacturing, if equipment fails and relies on message transmission and back-end commands, delays could worsen the situation. Similarly, in retail, integrating facial recognition with CRM for faster service may suffer if the back-end server handles all processing.
Moreover, the computational burden and data transmission costs of back-end systems are significant. The vision of IoT is to connect everything, but if all information is processed in the cloud, the server must be extremely powerful, leading to high infrastructure and operational costs. In such cases, edge computing becomes a better choice.
However, edge computing is not without its drawbacks. For instance, in environments like a car body, simultaneous operations at multiple points might cause interference, and preprocessed data at the edge may suffer distortion. Nevertheless, the development of IoT is not about choosing one over the other. Most systems operate in parallel, with edge computing used in areas requiring high real-time performance, while other parts remain centralized.
For Taiwan, centralized computing was never a major business opportunity. Traditionally, Taiwanese manufacturers focused on consumer products, and during the IoT boom, many still concentrated on first-tier equipment. The rise of edge computing aligns well with their product strategies and market conditions.
First, there's the computing chip. Previously, IoT terminal components needed to be low-power and compact, allowing devices to run efficiently in limited spaces. Now, with edge computing, some devices require more computing power—something most Fabless or IC designers can handle. Additionally, future IoT systems are likely to be vertical-specific, such as in manufacturing, healthcare, and transportation. These industries demand tailored integration, especially for on-site devices. Taiwanese manufacturers excel in rapid and flexible customization, which gives them an advantage in niche markets. However, these applications are also easy to replicate, so securing patents in specific areas is crucial for market stability.
The combination of AI and HI is the best answer.
One challenge for Taiwan is the lack of industrialization in AI. Despite having talented software engineers, past government and industry support has been lacking. Du Fu, who recently established an AI lab in Taiwan, noted that despite minimal government support, the Taiwan team consistently performed well at Microsoft’s annual developer conference, Build. This shows that Taiwan’s talent is strong, but it hasn’t always been recognized or supported.
Since 2018, the Ministry of Science and Technology has launched AI policies aimed at retaining software talent and advancing AI industrialization. With this, Taiwan can build a complete AIoT ecosystem, combining both hardware and software. For example, designing edge computing chips that are small, low-power, and powerful requires not only hardware expertise but also advanced algorithms. Software engineers must simplify complex models to allow efficient operation in low-power environments.
Although the integration of AI and IoT is still in early stages, the overall trend is clear. In June 2017, Alibaba founder Jack Ma suggested that the industry is moving from “Internet +†to “AI +,†meaning AI will combine with various fields to create added value. This aligns with the vertical application approach that IoT has long promoted. However, many experts argue that AI won't replace humans entirely. Sun Ki-kang, general manager of Microsoft Taiwan, emphasized that AI must be combined with Human Intelligence (HI) to create Super Intelligence (SI).
Today, AIoT aims to provide users with more intuitive, intelligent, and diverse options. However, logical decisions still require human input. In manufacturing and healthcare, for example, the Industrial IoT and AI have reached the perceptual level, moving away from cold manual instructions toward more human-centric messages. When a machine fails, the system can analyze past deep learning results and inform operators of the issue and possible causes. Staff can then use the system's suggestions along with their own expertise to resolve the problem. If immediate action is required, the system can alert staff via voice, enabling quick response.
In healthcare, AI can analyze medical data or images, but real diagnosis and treatment still depend on professional doctors. This is how AI and HI work together.
After integrating AI, IoT will accelerate its application development. From an industrial architecture perspective, the advantages Taiwanese manufacturers once had in consumer products can now extend to terminal components and networked devices in IoT systems. These products require some level of customization, presenting a challenge and an opportunity for Taiwanese manufacturers. They must invest more resources to gain expertise in specific fields, enhancing product value and breaking free from the low-value OEM model. This positions them to capture more business opportunities in the growing AIoT market.
September 24, 2025