原创 自动化技术在食品加工中有着广泛的应用。下面翻译了记者Jimmy Carroll对Harley Green的一段采访。

自动化技术在食品加工中有着广泛的应用。下面翻译了记者Jimmy Carroll对Harley Green的一段采访。

Jimmy Carroll: Hi, Everyone. My name is Jimmy Carroll. I'm an A3 contributing editor and VP of operations at Tech B2B Marketing. I'm here today with Harley Green, director of business development at Soft Robotics, and we're here to discuss automation and food processing. Thanks very much for taking the time to meet with me, Harley.

Harley Green: Thank you, Jimmy.

Jimmy: Of course. So, across several different industries, adoption of automation technologies like robots and machine vision and AI continues to grow. Among these areas is the food industry, which is seeing significant growth in terms of robot sales. A3 figures show that that robot sales in food and consumer goods grew 13% year-over-year in Q2, while the International Federation of Robotics, the IFR, shows that the food and beverage industry installed a record number of robots in 2021 — 25% more than were installed in 2020. Obviously, events such as COVID and the workforce shortage play into this, but it seems like the food industry was probably primed for growth in terms of automation regardless. My question to you is, do you agree? And if so, can you talk a bit about the different ways you've seen food processing companies deploy automation?

Harley: Yeah, 100%. The focus recently has been around, “Oh, my God, we can't find labor” or “we can't retain labor. How do we reduce our reliance on labor?” But to your point, I think before that we had already seen some companies in the food industry starting to automate. The baked goods space in particular has been using robotics for years. A lot of robots are feeding flow wrappers — such as putting cereal bars and all sorts of products onto a conveyor to feed the system. Outside of that, if you look at some of the proteins, there's some pretty cool solutions out there in the primal cutting and sub primal cutting space where you might attach, let's just say, a dangerous tool — a blade on the end of a robot — that will make cuts on, say, a piece of beef or pork. And we've seen other areas, whether it be primary packaging or secondary packaging, or palletizing, where either a lot of people are required to do mundane tasks or companies are looking to reduce some of the ergonomic strain issues that might come along with these jobs. You know, lifting heavy boxes, moving them over their heads, and things like that. So, we've definitely seen a lot of growth in that space.

Some companies choose to focus downstream on palletizing first and then move further upstream. But one of the key goals that most food manufacturers have is that they're looking for something capable of replacing the value that people bring to an operation. Here, you start getting into, okay, now I really need human hand-like dexterity or the meaningful decision making that people do while in a chaotic environment, for instance.

Jimmy: What are some of the primary challenges that food processing companies run into when it comes to deploying automation, and how are they addressing or how can they address these issues?

Harley: First and foremost, what we hear from most customers is floor space, right? It's at a premium. A lot of food manufacturing facilities have been around for a long time and weren’t designed or built with automation in mind. For the most part, you can sort of put people shoulder to shoulder in a facility. This was a problem during the pandemic, but otherwise people don’t take up too much space. But robots and automation in general need a little bit more structure. And floor space comes at a premium. So, I would say the number one challenge that most food processing companies run into when they start looking to automate is they don't have the floor space needed. Now, that also leads to a great opportunity for certain manufacturers of technology and equipment to then come up with solutions in the allotted space.

Another problem is all the different food products. Unless they're stamped out, sort of ready to eat or convenience food–type products, there is a lot of variability to them. No two apples are the same. No two avocados are exactly the same. So, there's this natural variability that has to be accounted for that you don't typically find with automation. Everything in automation is generally very structured. Finding a way to bridge the gap between the variability that you see in food and what robots have currently been able to do is a factor.

And then there’s always food safety. That is the number one concern outside of floor space with food manufacturers trying to automate. They want to make sure that they are not introducing any additional foreign material into their products or their packaging. That would obviously be a big problem. They want to make sure that all the people providing solutions into this space are able to check the boxes indicating that these are hygienic systems — parts can be easily cleaned, components are food grade, and things of that nature. It all revolves around food safety.

Jimmy: One of the things that we see is that as processes evolve, machine vision vendors need to evolve their offerings to meet evolving needs. That means that a lot of different types of technologies are used in food processing today. What are some of the trends in food processing that have pushed the vendor community to innovate further? I'm thinking of things like IP69K protection, 3D developments, multispectral, hyperspectral, anything like that.

Harley: A lot of what you just touched on goes back to my previous comment. In general, the technology has to be food safe and IP69K protection is a critical piece in making sure that equipment manufacturers look at that from the ground up and have that be a requirement in their solution.

The food space has historically been very slow to adopt automation. Therefore, a lot of the component manufacturers haven't necessarily come up with IP69K-rated, food grade or hygienic solutions, so, there's a little bit of a gap there. I do think that we're seeing some trends move toward that as component manufacturers start to see some other markets slowing down. But food, as you mentioned, is growing at an extreme pace for automation. The end users, the food manufacturers, need to be able to check the box that says this is a hygienic, indoor IP69K-rated solution. Now I can move on to the next step, which is usually, “How can I now handle my production?”

People are really good at looking down at a pile of chicken drumsticks coming down a production line and being able to pick them out one by one and put them into a tray. Well, robots are not so great at that. But we’re seeing new developments in 3D technology and different imagers being able to see and handle different scenarios. We’ve gone from having to have everything perfectly presented to a robot — the same product, the same shape, the same size and weight every single time — to robots that are able to handle piles of product moving down a conveyor just like people would.

One thing that I always look at is the cost for a company to have a food recall. Right? And one thing that we always try to do is make sure that anything that goes into this food space is not going to cause any issues. It's not going to be a cause for concern for the food manufacturer and is ultimately going to make them have higher throughput and output at the end of the day.

Jimmy: When it comes to deploying these different technologies, food processing companies have options. They can use an internal integrator or have a third-party integrator come in and design, specify, and integrate a system. But one trend that I've observed over recent years, and this applies to soft robotics, of course, is the increase in the availability of machine vision or robotic systems that target specific needs. Your company, Soft Robotics, is an example of a company that offers such a system. Why do you think these so-called turnkey systems have become more popular in food processing as of late?

Harley: It all goes back to labor, right? And trying to close the gap between the labor needs of each of these manufacturers and the value that humans bring to those environments. Speed to deployment, for instance. How can end users take advantage of this technology today rather than tomorrow to satisfy some of their needs? And we’ve seen a lot of advancements in machine vision and AI solving some very challenging issues, like I have so much variability in my product. How can I take one solution and one approach to solving this? And the reality is, you know, maybe there isn't just one approach. Maybe you really need some additional horsepower to be able to handle all that flexibility.

The food space in general has historically relied on humancentric applications. You've really got to capture the value of what people bring to these environments and apply that to robots. People are really good at making adjustments on the fly. If you're a food manufacturer and you're running a flex line, meaning that you're going to have different products running on the same line at different days or different weeks, you need assurance that the system is going to be able to switch over quickly and easily with minimal, if any, disruption to manufacturing.

As a nation, we continue to demand more food. Food consumption has gone up, and manufacturers don't want to have any negative impact on the supply chain. So, the top drivers of these turnkey systems are speed of deployment, trying to solve the labor gap and trying to keep up with consumer demand.

Jimmy: Last question for you, Harley. What is one thing that any food processing company looking to implement automation or say, upgrade an existing system, should really do before starting that process?

Harley: We work directly with end users. And I think the number one value that they can bring early on to this process is truly understanding their entire process and their problems. And I would say that most automation projects are rarely just one dimensional. I like to say, look at the upstream processes, look at the downstream processes, and really know the problems so that we can help you right-size your implementation of automation.

It’s best for end users to be honest about their processes and what they want to achieve, what their goals are, so when we all come together at the end of this, everyone's going to be better aligned with the end solution, and it's going to meet the expectations of the customer. Do the upfront homework.

Some food manufacturers haven't even thought about any of this because the business has just been so chaotic. But really understand your process, understand your goals, and make sure that you bring that sort of honest insight into the project. And think about the future as well. You know, don't just think about, “How do I keep up with my production today?” Think about whether you plan on increasing production for the future. You don't want to necessarily box yourself in with the solution.

Jimmy: Harley, thank you for taking the time today to speak with me. I appreciate it. If anyone has any questions or comments for A3 on this topic, I'd encourage you to visit www.automate.org. Otherwise, thanks for checking this out.

Harley: Thanks, Jimmy.

吉米·卡罗尔:大家好。我叫吉米·卡罗。我是A3特约编辑和技术B2B营销的运营副总裁。今天我和软式机器人公司的业务开发总监哈利·格林在一起，我们来讨论自动化和食品加工。非常感谢你抽时间见我，哈利。

哈利·格林:谢谢你，吉米。

吉米:当然。因此，在几个不同的行业，机器人、机器视觉和人工智能等自动化技术的采用持续增长。在这些领域中，食品行业的机器人销量正在显著增长。A3数据显示，第二季度食品和消费品领域的机器人销量同比增长13%，而国际机器人联合会(IFR)显示，食品和饮料行业在2021年安装了创纪录的机器人数量，比2020年增加了25%。显然，COVID和劳动力短缺等事件在其中发挥了作用，但食品行业似乎已经为自动化方面的增长做好了准备。我的问题是，你同意吗?如果是这样，你能谈谈你看到的食品加工公司部署自动化的不同方式吗?

哈利:是的，百分之百。最近的焦点是，“哦，天哪，我们找不到劳动力了”或“我们留不住劳动力了”。我们如何减少对劳动力的依赖?”但就你的观点而言，我认为在此之前，我们已经看到食品行业的一些公司开始实现自动化。烘焙食品领域多年来一直在使用机器人技术。很多机器人都在给流包装送料——比如把谷物棒和各种各样的产品放到传送带上，给系统送料。除此之外，如果你观察一些蛋白质，在原始切割和次原始切割空间里有一些很酷的解决方案，比如说，你可以附加一个危险的工具——机器人末端的刀片——它会在一块牛肉或猪肉上进行切割。我们还看到了其他领域，无论是初级包装还是二级包装，或者是码垛，要么是很多人被要求做平凡的工作，要么是公司希望减少这些工作可能带来的人体工程学压力问题。举重箱子，把箱子举过头顶，诸如此类的事情。所以，我们确实看到了这个领域的巨大增长。

一些公司选择先将重点放在下游的码垛上，然后再进一步向上游发展。但大多数食品制造商的一个关键目标是，他们正在寻找一种能够取代人们给经营带来的价值的东西。在这里，你开始进入，好吧，现在我真的需要像人手一样的灵巧，或者有意义的决策，比如人们在混乱的环境中做出的决定。

吉米:在部署自动化时，食品加工公司遇到的一些主要挑战是什么?他们是如何解决或如何解决这些问题的?

Harley:首先，我们从大多数客户那里听到的是占地面积，对吗?这是一种溢价。许多食品生产设施已经存在很长时间了，在设计或建造时并没有考虑到自动化。在大多数情况下，你可以把人们肩并肩地放在一个设施里。这在大流行期间是一个问题，但除此之外，人们不会占用太多空间。但是机器人和自动化通常需要更多的结构。楼面面积非常昂贵。所以，我想说，大多数食品加工公司在开始寻求自动化时遇到的第一个挑战是他们没有所需的地板空间。现在，这也为某些技术和设备制造商提供了一个很好的机会，然后在分配的空间内提出解决方案。

另一个问题是各种不同的食品。除非它们被剔除，像即食食品或方便食品一类的产品，否则它们有很多可变性。没有两个苹果是一样的。没有两个牛油果是完全一样的。所以，必须考虑到这种自然的可变性这在自动化中是不常见的。自动化中的一切通常都是非常结构化的。找到一种方法来弥合你在食物中看到的可变性和机器人目前能够做到的之间的差距是一个因素。

还有就是食品安全问题。这是食品制造商试图自动化生产时，除了占地面积之外的头号问题。他们想要确保他们没有将任何额外的外来物质引入他们的产品或包装。这显然是个大问题。他们希望确保所有为这个领域提供解决方案的人都能够在方框中勾选，表明这些是卫生系统——部件可以轻松清洁，组件是食品级的，以及类似的性质。这一切都围绕着食品安全。

Jimmy:我们看到的一件事是，随着流程的发展，机器视觉供应商需要发展他们的产品，以满足不断变化的需求。这意味着在今天的食品加工中使用了很多不同类型的技术。食品加工领域有哪些趋势推动了供应商社区的进一步创新?我想的是IP69K防护，3D开发，多光谱，高光谱，诸如此类的东西。

哈利:你刚才提到的很多内容都可以追溯到我之前的评论。一般来说，技术必须是食品安全的，IP69K防护是确保设备制造商从头开始考虑并将其作为解决方案的一个要求的关键部分。

从历史上看，食品行业采用自动化的速度非常缓慢。因此，很多部件制造商不一定能提出ip69k级、食品级或卫生的解决方案，所以，这里有一点差距。我确实认为，随着零部件制造商开始看到其他一些市场放缓，我们正在看到一些趋势朝着这个方向发展。但是，正如你所提到的，食品正在以自动化的极端速度增长。最终用户，也就是食品制造商，需要能够勾选这是一种卫生的，室内ip69k级的解决方案。现在我可以进入下一个步骤，通常是“我现在如何处理我的生产?”

人们真的很擅长低头看着生产线上一堆鸡腿，并能够一个一个地把它们挑出来，放进托盘里。机器人在这方面不太擅长。但我们看到3D技术的新发展，不同的成像仪能够看到和处理不同的场景。我们已经从必须把所有东西完美地呈现给机器人——每次都是相同的产品，相同的形状，相同的大小和重量——到能够像人一样处理一堆堆产品从传送带上移动的机器人。

我经常关注的一件事是公司进行食品召回的成本。对吧?我们一直试图做的一件事是确保进入这个食物空间的任何东西都不会引起任何问题。这不会成为食品制造商担心的原因，最终会使他们在一天结束时有更高的吞吐量和产量。

吉米:当涉及到使用这些不同的技术时，食品加工公司有多种选择。他们可以使用内部集成商，也可以让第三方集成商来设计、指定和集成系统。但我最近几年观察到的一个趋势，当然也适用于软机器人，是针对特定需求的机器视觉或机器人系统的可用性的增加。您的公司Soft Robotics就是提供这种系统的公司的一个例子。你认为为什么这些所谓的交钥匙系统在食品加工中越来越受欢迎?

哈利:这一切都回到了分娩，对吧?试图缩小每个制造商的劳动力需求和人类给环境带来的价值之间的差距。例如，快速部署。终端用户如何在今天而不是明天利用这项技术来满足他们的一些需求?我们在机器视觉和人工智能方面看到了很多进步，解决了一些非常具有挑战性的问题，比如我的产品有很多可变性。我如何采取一种解决方案和方法来解决这个问题?而现实是，你知道，也许并不是只有一种方法。也许你真的需要一些额外的马力来处理所有的灵活性。

在历史上，食品空间总体上依赖于以人为中心的应用。你必须抓住人们给这些环境带来的价值，并将其应用到机器人身上。人们真的很擅长在飞行中做出调整。如果你是一家食品制造商，你正在运行一条灵活的生产线，这意味着你将有不同的产品在不同的日子或不同的星期在同一条生产线上运行，你需要确保系统能够快速轻松地切换，对生产造成最小的干扰。

作为一个国家，我们继续需要更多的食物。食品消费量上升了，制造商不希望对供应链产生任何负面影响。因此，这些交钥匙系统的主要驱动因素是部署速度，试图解决劳动力缺口，并努力跟上消费者的需求。

吉米:最后一个问题，哈利。任何食品加工公司想要实现自动化，或者升级现有的系统，在开始这个过程之前应该做什么?

哈利:我们直接与终端用户合作。我认为他们在这个过程中最重要的价值是真正理解他们的整个过程和他们的问题。我想说的是，大多数自动化项目很少只是一维的。我想说的是，看看上游流程，看看下游流程，真正了解问题，这样我们就可以帮助您正确地实现自动化。

对于最终用户来说，最好是诚实地说出他们的流程，他们想要实现什么，他们的目标是什么，所以当我们最终聚在一起时，每个人都将更好地与最终解决方案保持一致，并且它将满足客户的期望。提前做好功课。

一些食品制造商甚至没有考虑过这些问题，因为他们的行业太混乱了。但是要真正理解你的过程，理解你的目标，并确保你把那种诚实的洞察力带入到项目中。也要想想未来。你知道，不要只是想，“我今天如何跟上我的生产?”考虑一下你是否计划在未来增加产量。你不想用解决方案束缚自己。

吉米:哈利，谢谢你今天抽出时间和我谈话。我很感激。如果任何人对A3这个主题有任何问题或意见，我鼓励你访问www.automate.org。否则，谢谢你看这篇文章。

哈利:谢谢，吉米。