|Abstract: ||近年來機器人應用盛行，許多產業也開始投入相關應用的開發，對於餐飲業需要大量人力，更是主要結合重點。因此我們希望在送餐點這部份可以由機器人來取代，原因在於可以減輕人力在移動中所花費的時間，但在現有的送餐機器人中，客人只能自己動手拿取餐點，為了讓送餐系統更人性化，所以我們結合移動機器人和機械手臂，最後透過手臂將餐點放置定點桌上。 這份論文中，我們利用機器人對於送餐系統提出了一組有效的方法，我們將系統分成四個階段。在第一階段中，由於送餐環境大多都在室內，因此機器人需要對環境建構地圖，利用雷射感測器使現實場景回傳周圍物體位置來建構場景。第二階段，利用第一階段生成的地圖做為室內導航的依據，由地圖的起始點到終點算出最短路徑，我們稱為全域規劃。在第三階段，針對在路徑上的障礙物避障，立即的躲避以及重新規劃路徑，我們稱為區域規劃。最後一個階段為機械手臂的運作，機器人必須知道桌面的高度才能將餐點放置上去，因此使用圖標作為桌邊高度，利用影像處理將圖標辨識得知其位置，根據位置操作手臂旋轉角度。 有別於現有的送餐機器人，我們新增的三、四階段讓機器人更加接近人性服務客人，除了透過激光雷達來感知現實周圍的情況，我們還利用了電腦視覺處裡來辨別桌子的高度近來控制手臂旋轉位置。 我們的系統不僅提供了移動機器人結合機械手臂的概念，更在送餐系統上貼切的提供服務，客人不需要花而外的力氣自己取餐，而服務員也不需要再把時間花在重復性高的動作上，達到節省人力時間之共用。相信藉由我們的研究發展，對於國內的機器人產業及餐飲業必能有所助益。|
In recent years, robot applications have prevailed, and many industries have begun to invest in the development of related applications. For the catering industry, a large amount of manpower is required, and the main focus is on combining. Therefore, we hope that this part of the meal can be replaced by a robot, because it can reduce the time spent by manpower on the move, but in the existing meal delivery robot, the guests can only take their own meals, in order to let The food delivery system is more user-friendly, so we combine the mobile robot and the robotic arm, and finally place the meal on the fixed table through the arm.In this paper, we use robots to propose an effective method for the meal delivery system. We divide the system into four phases. In the first stage, since most of the food feeding environment is indoors, the robot needs to construct a map for the environment, and use a laser sensor to make the real scene back to the surrounding object to construct the scene. In the second stage, the map generated in the first stage is used as the basis for indoor navigation. The shortest path is calculated from the starting point to the end point of the map. We call it the global planning. In the third phase, we are called regional planning for obstacle avoidance, immediate avoidance, and re-planning of obstacles on the path. The last stage is the operation of the robot arm. The robot must know the height of the table to place the meal. Therefore, the icon is used as the height of the table edge, and the image is used to identify the position of the icon and operate the arm rotation angle according to the position.Different from the existing feeding robots, our new three or four stages make the robots closer to the human service guests. In addition to perceiving the reality around the laser, we also use the computer vision to identify the height of the table. Control the position of the arm rotation.Our system not only provides the concept of a mobile robot combined with a robotic arm, but also provides a service on the food delivery system. Guests do not need to spend their own time to take their own meals, and the waiter does not need to spend time on repetitiveness. In terms of the action, the sharing of labor saving time is achieved. It is believed that with the development of our research, it will be helpful for the domestic robot industry and the catering industry.