隨著智慧化時代正式來臨,人工智慧與物聯網技術已經被廣泛應用於各個領域,人工智慧更經常被應用於智慧機器人,國內外也出現很多智慧機器人的實作競賽。其中智慧機器人競速一直是國內外經常比賽的項目之一,常見的比賽類型有以視覺感測器辨識賽道及號誌使智慧機器人完成賽道。這類的競賽通常會使用樹莓派當作智慧機器人的運算平台並且進行機器視覺辨識處理。本論文以AIoT智慧機器人競速賽為例,分析在樹莓派平台中有採用機器人作業系統與沒有採用機器人作業系統的競速效能與辨識穩定度差異。經過AIoT智慧機器人競速賽的實驗測試後,在樹莓派平台中有採用機器人作業系統比沒有採用機器人作業系統的競速效能提升70%以上及辨識穩定度提升160%以上。
With the coming of the era of intelligence, Artificial Intelligence and Internet of Things technologies have been widely used in various fields. Artificial Intelligence is more often applied to smart robots, and there are many smart robot implementation competitions. Among them, smart robot racing has always been one of the frequent competitions. Common types of competitions include visual sensors to identify the track and signs to make the smart robot complete the track. This type of competition usually uses the Raspberry Pi as a computing platform for smart robots and performs machine vision recognition processing. Taking the AIoT smart robot racing competition as an example, the paper analyzes the difference in racing performance and recognition stability between the robot operating system and the robot operating system not used in the Raspberry Pi platform. After the experimental test of the AIoT smart robot racing competition, the robot operating system on the Raspberry Pi platform has increased by more than 70% and the recognition stability increased by more than 160% compared to the racing performance without the robot operating system.