AI物聯網大合體 智慧應用再進化 (2018-08-02 IEK產業情報網 )

The ITRI IEK expects an accelerating convergence of AI and IoT, into AioT, in 2018, and this will create a plethora of smart applications. As the key AI technologies continue to progress, the industry is faced with a challenge in the choice of technology roadmaps and the identification of killer applications in digital transformation.
The vision is to make AI ubiquitous and easy to use. The evolution of AI’s role from perception and awareness to decision making paves the way for the next stage of AI development.
The majority of existing AI applications are on the cloud because only large computing centers can support the resources required for deep learning. However, the prevalence of AI will eventually shift to mobile and edge devices. The two technical hurdles for the architecture of edge computing are low power consumption chips and lightweight algorithms suitable for edge devices. The global semiconductor industry has been focusing on the development of low power consumption chips for some time. Various solutions have been put in place, each with its own scenarios, pros and cons. Created for edge devices, lightweight algorithms are the streamlined versions of massive algorithms via compression technology. International tech giants and start-ups are all developing the techniques and services required.
AI has come a long way in terms of recognition and perception. The end game is to make AI able to think like humans. This means faster learning, less data input, the ability to solve more complex decision-making problems, and autonomous learning of the required skills. Democratization of AI is also a key focus for the next stage, to bring AI to more individuals and companies.
IEK expects AI to gradually mature and penetrate all industries. The fundamental revolution in the bottom layer of AI will enhance its intelligence and capability and help us to solve more problems. In the development of AI applications everywhere, companies in Taiwan should also stay on top of AI transformations. Every wave of changes in AI technology promises to bring significant improvements to a firm’s competitiveness.
IEK has identified three trends in smart home. First of all, Amazon’s Echo emerged in 2015 as a disruptive force with its AI-powered voice assistant reigniting the global market for smart homes. In the second half of 2017, the market for smart speakers entered an explosive growth period. Second, the value chain for AI-enabled voice services continues to take shape in both English-speaking and Chinese-speaking regions. Third, the battle for home robots has started, and this is shifting the competitive landscape for leading players in the smart home sector. Tablet makers, white goods manufacturers, surveillance device/service suppliers, mobile service operators and automobile companies are all entering the domestic robot market from their home turfs, by focusing on family care and automated control.
AI-powered commerce is expanding, driven by AI and IoT. This will open up the demand for a variety of AioT hardware and software solutions, such as delivery drones, driverless taxis, unmanned stores, facial recognition payments, smart signage, smart dining tables, smart shelves, comforting or social robots, business support robots, warehousing and logistics robots and cargo inspection robots. AI and the platform serve as the brain of AioT, and are critical to the smooth operation of the systems.
IEK believes that AI-enabled service robot systems and platforms will develop in the following seven directions: cobots, voice assistants, machine vision, mobility platforms, applications centered on users’ demands, self-learning algorithms, and open hardware/software architecture for robots. Meanwhile, AI is also driving the transformation in smart workflows, products and services. The concept of the Digital Twin is shaping up as a result of the convergence of AI, AR, IoT and PaaS.
Going forward, AI application platforms, AI algorithms and AI sensors will essentially become one. Companies in Taiwan may respond by centering on AI application systems, platforms and architectures in order to get a grip of relevant AI algorithms. Taiwanese players may seek to make inroads by utilizing their advantage in the components for AioT edge devices and IoT equipment (such as sensor chips, optical modules, touch panels and gateways). Any enhancement in the functionality of key components will assist Taiwanese companies in the development of software modules and integration with hardware solutions. The offering of high value-added components for AIoT machines could be a way to boost Taiwan’s competitiveness in AIoT among Taiwanese players.
As the equation “AI+IR=AIR” implies, the value of AI is manifested with the advent and advancement of robots. With the rapid development of computing power, broadband and mobile devices, the AI ecosystem continues to expand. The improvement of computing power is also accelerating recognition technologies which, in conjunction with IoT, can collect a massive amount of real-time data (“Big Data”). The evolution and growth of the AI ecosystem will bring service robots into our everyday life. By then, robots will be connected to the network and IoT devices and will be communicating with us via an AI engine. Before long, virtually all the product lines from global IT leaders are going to be associated with AI in one way or another.
比特幣BTC 12861.90 38.21 0.30%
以太幣ETH 396.93 4.74 1.21%
瑞波幣XRP 0.258505 0.01 2.74%
比特幣現金BCH 263.29 4.78 1.85%
萊特幣LTC 54.84 1.79 3.37%
卡達幣ADA 0.107187 0.00 0.94%
波場幣TRX 0.026866 0.00 0.54%
恆星幣XLM 0.085855 0.00 2.48%
行動版 電腦版
系統合作: 精誠資訊股份有限公司
資訊提供: 精誠資訊股份有限公司
資料來源: 台灣證券交易所, 櫃買中心, 台灣期貨交易所
依證券主管機關規定,使用本網站股票、期貨等金融報價資訊之會員,務請詳細閱讀「資訊用戶權益暨使用同意聲明書」並建議會員使用本網站資訊, 在金融和投資等方面,能具有足夠知識及經驗以判斷投資的價值與風險,同時會員也同意本網站所提供之金融資訊, 係供參考,不能做為投資交易之依據;若引以進行交易時,仍應透過一般合法交易管道,並自行判斷市場價格與風險。
請遵守台灣證券交易所『交易資訊使用管理辦法』等交易資訊管理相關規定本資料僅供參考,所有資料以台灣證券交易所、櫃買中心公告為準。 因網路傳輸問題造成之資料更新延誤,精誠資訊不負交易損失責任。