【弗图尼尔】机器人抓取包裹​

Bin picking – i.e. robotic detecting andpicking of randomly arranged objects from within a bin on the basis of CADmodels – is a common type of application we deal with even in challenging industrial environments. Our BinPicking Studio is an all-in-one picking solution for robotic integrators, who do notneed to know anything about computer vision to set it up. Its robust featuresmade it possible to enter basically allindustrial sectors – the automotive, manufacturing, logistics, and others.

自动抓取系统——即基于CAD模型的机器人检测和选取箱内随机排列物体——是我们在具有挑战性的工业环境中处理的常见应用类型。我们的自动抓取系统为机器人集成商提供一个一体化的解决方案，机器人集成商不需要掌握任何计算机视觉（知识）来设置它。这一强大的特点使它有可能进入几乎所有的工业部门——汽车、制造、物流及其他。

However, we gradually started to receiverequests from our customers for a solution that would enable them to pick objects of irregular shapes and in randompositions within the bin. We set to work on algorithms that would enable usto pick these kinds of objects and after some time of intensive endeavor, our AI-driven solutions first saw the lightof day. Thanks to sophisticated algorithms, they enabled us to localize andpick objects of different sizes, various textures, and arbitrary shapesincluding deformations. What’s more, the objects do not need to be stacked inordered patterns but can be placed absolutely randomly and the robot is stillable to recognize and pick them. Our AI-powered solutions enabled by a combinationof precise 3D vision and advanced AI algorithms showed the world the immenseopportunities for automation of materialhandling, manufacturing, logistics, and many more.

然而，我们逐渐开始接受客户提出的解决方案，让他们能够在箱内挑选不规则形状和随机位置的物体。我们开始研究使我们能够选择这类物体的算法，经过一段时间的集中努力，我们的人工智能驱动的解决方案第一次看到了曙光。多亏了复杂的算法，它们使我们能够定位和选择不同大小、不同纹理与包括变形在内的任意形状的物体。更重要的是，物体不需要以（整齐）有序的方式堆叠，可以绝对随机地放置，机器人仍然能够识别和挑选它们。我们的人工智能解决方案结合了精确的3D视觉和先进的人工智能算法，向世界展示了在材料处理、制造、物流等方面实现自动化的巨大机遇。

And it was logistics that showed us therewas one last gap left to cover all kinds of objects and applications – and thisresided in the fact that goods are not always transported in solid boxes butsome of them come in bags. Thesepose a great challenge for automated sorting as they do not keep their shape –they are flexible, wrinkled, easilydeformable, and thus very difficult to localize. We decided to utilize ourexperience, knowledge, and skills to pushthe performance of our AI-powered solutions even further and “teach” themdetect and pick bags.

正是物流让我们看到，还有最后一个空白，可以涵盖所有类型的对象和应用程序—这是因为货物并不总是装在坚固的盒子里运输，而是有些装在袋子里。这给自动排序带来了巨大的挑战，因为它们不能保持它们的形状—它们是灵活的，皱巴巴的，容易变形的，因此很难定位。我们决定利用我们的经验、知识和技能进一步提升我们的人工智能解决方案的性能，并“教”他们检测和挑选包。

The recognition of boundaries betweenbags that are chaotically placed in a container can often be difficult even forhumans. The task gets even more complicated for transparent bags. Difficult – but possible. We developed a networkthat is able to recognize bags, sometimes better than we are. They may be overlappingand blocking each other from being picked but our system can handle this withease and won’t allow the robot to grabfully or partly covered bags.

即使是对人类来说，要识别被杂乱地放置在容器中的袋子之间的界限也是很困难的。对于透明塑料袋来说，这个任务更加复杂。（这个任务）困难但有可能（完成）。我们开发了一个能够识别袋子的网络，（这个网络）有时比我们自己识别的还要好。它们可能是重叠的，并且相互阻碍被选中，但我们的系统可以轻松处理这个问题，不会允许机器人抓取全部或部分覆盖的袋子。

Another common difficulty arises from thenature of the material bags are made of – they are full of folds and wrinkleswhich often causes them to fall off the gripper just after being successfullydetected and lifted. We recommend using a vacuumgripper with feedback to prevent such failures.

另一个常见的困难来自于袋子的材质——它们充满折痕和褶皱，这导致它们经常在被成功地发现和举起后从夹持器上掉下来。我们建议使用带反馈的真空夹持器，以防止此类故障。