BackLooking for Collaborative Robot Applications with the Right Mindset
Table of Contents
How to look at application speed?
It is important to understand what cobots are designed for and what their strengths and weaknesses are because of their intended use.
How fast is a cobot?
Cobot stands for collaborative robot, an industrial robot designed to work with humans. This is crucial, because it adds different safety requirements and limits on the cobot to protect the humans’ well being and therefore causes payload and speed restrictions for the automated application.
You should have this in mind when looking at applications as it’s not that the robot couldn’t move faster, but instead it shouldn’t move faster. Or better, be able to stop fast enough at any given time to qualify as a cobot. If you require extreme speeds to catch up with your target cycle time, you may look at the wrong process to automate with a cobot. Of course, there are solutions how processes can be automated smartly to buy more time, but we are looking here at the right mindset for the cobot application, so you better keep looking for a better collaborative application.
The cobots’ competitor
In Thailand, there are a lot of manual processes done by operators, so it’s important to understand that these operators are generally equipped with two arms. In most applications based on the budget and the Thai salary level, the AUBO is expected to compete with a two-armed human, so this needs to be factored in when considering the automation solution.
If there is a double shift system in place, budget-wise we can compete with two robot arms as a general rule and we are much closer to make it a more reliable and consistent automated solution, which saves costs at the same time.
Recommendations & Best Practice
A good practice is to look at the operator as a theoretical humanoid robot. It has two arms and quite a number of sensors like vision, touch and hearing, powered by a very fast and powerful CPU, the brain. As a rule of thumb, an implemented robot arm is in direct comparison to a human arm at about 70% of the human speed based on additional time required for process control like sensor tests or alignment moves and limitations due to safety requirements.
Background
The required safety for collaborative robots is described in the ISO/TS 15066:2016 which defines fundamental requirements regarding cobot safety. The main points described in it are the allowed forces when getting in contact with humans or more specifically, what forces in what body regions and in which time/distance the robot has to stop when a contact is detected or a safety system is triggered.
Make a cobot faster?
This is the reason why cobots are limited in what maximum speed they can achieve in collaborative mode. However with the use of additional safety measures and a risk assessment, it is possible to run them faster in the non-collaborative mode while still taking advantage of the easy setup and programming.
How to look at application complexity?
Too complicated applications at the wrong time is one of the most common reasons for robot automation to be considered unsuccessful.
What are the right applications to start with?
We see over and over wrong applications being selected, especially when making the first steps with cobots; or robots in general for that matter. Complex applications come at a price, which may be just long development times or additional technology needed to automate the task at hand.
Production people (no disrespect, I just have no other short form of categorizing) love to throw their biggest quality challenges or RnD products, which tend to have small quantities and are often rather complex, at the first-time integration of robot arms. This is generalizing and there are some significant exceptions to this statement, but there are usually reasons why the challenges are not easy to overcome.
Choosing such applications to begin the transformation in the production is just too often a cause for frustration down the road, so it should be avoided. Simpler is better, at least for the beginning, since it is also very motivating to get the first steps done rather quickly and create acceptance of the new “colleague” in the factory. Later, this will lead to addressing the production headaches with a solid foundation built up by simpler, successfully finished implementations.
How to start?
In many cases we even recommend starting with the first project, which doesn’t achieve the standard ROI requirements. Simpler means in most cases that little additional investment is required. At the end of the day, it’s a fully programmable robot arm which can be quickly adjusted to more cost efficient tasks since this is one of the main benefits of cobots. You don’t need a forklift to move it or a massive foundation to support it so you can absolutely make use of the flexibility to move and quickly set it up with a different task at any time.
One of the main rules is to do the right thing at the right time, so choose the application which fits your current robot or application skill level and the confidence to implement fast.
Recommendations & Best Practice
My main recommendation to evaluate the complexity of an application is to close your eyes, put one arm behind your back and try to imagine executing the task. This is what often boils it down, because you will notice how much information you are actually missing; like if the machine I try to unload is missing, is there any inspection which is done by the operator just by looking at the product and so on.
It will give you a way to start and make the best choice or allow you to rate the vast number of possible applications you have to automate to select the best one to start or move on after the last successful implementation.
Background
There is a lot of excitement when starting with AUBO and the first robot application, because you don’t have to be an engineer to do the programming for the majority of applications. However, this often leads to feeling so confident that too complex tasks are selected for the actual level of solution experience.
A Beginners’ Trap
One of the most underrated side effects from cobots is that you can skip a lot of parts in your journey compared to let’s say 1000 kg robot arms. The, for safety purposes required, offline simulations of the applications can often be skipped and the safe programming with the real cobot and the teach mode, which allows you to move the robot arm by hand with a push of a button, quickly sparks euphoria as the focus shifts from “How do I make the robot do it?” to “What do I make the robot do?”, which is the actual manufacturing process development.
There is often such quick success on the first steps which customers didn’t expect, due to a presentation in the meeting room for example, that the ideas run wild and everybody tries to get rid of their worst production issues, which statistically are applications which require highly complex solutions, with the first cobot deployment in their company.
How to look at maximized utilization?
With maximized utilization we are looking at the OEE, the overall equipment efficiency. Specifically at the part of machine availability and its utilization. Strictly speaking about OEE, we would look at availability issues of a machine or production line. So let’s assume a scenario, in which we want to maximize the robot run time and its impact on a station’s OEE.
Human Working Hours
We are looking here into the possibility of a robot loading and unloading a production machine and we want to maximize the utilization based on a single shift setup. Generally speaking, whenever humans are working, there are breaks to take which the robot doesn’t have to take. An increase of utilization can be achieved, when the robot is running the full 8 hours, although in Thailand, 10 hour or 12 hour shifts are more common.
Our investment is now calculated over 9 hours utilization time instead of 7.5 hours for this example. We keep running through the one-hour lunch time break as well as the breaks before and after of 15 minutes each. However, this would only be possible if the robot application is not interrupted by other means like not having sufficient material coming into the process or sufficient capacity for material leaving the process.
Material supply in an automated production line
One of the more interesting examples, which often comes to mind, is the robot being in line with a production machine. Like an inspection, where the material passes through the station continuously and the robot just executes its task. This is the optimal utilization considering the material flow isn’t interrupted, because the material generating machine or process is run continuously throughout the break time as well.
Material supply with buffers
Another scenario to consider is a station with a material buffer before and behind the AUBO. This is very common, so the buffers for material in and out need to be sufficiently stocked, so the robot is continuously supplied. This is a factor not to be underestimated when looking at the best utilization, because often when the cycle time is small, there simply isn’t enough buffer space for the robot arm to execute it’s task throughout breaks.
OT rate of a cobot
An additional common concept for increase of utilization is extending the production time for the robot process after the single shift ends. If a high number of robot automation is already achieved, additional production hours beyond the single shift can be established by sending the robot in overtime. This can happen at a much cheaper rate compared to a human equivalent, especially with the low power consumption of cobots.
Based on that, a two-shift production with the robot being able to run through break times is the highest utilization which can be achieved.
Recommendations & Best Practice
Look for applications, in which the robot does not rely on human interaction for material transport by either being integrated in line with other systems or having an independent automatic material supply. This can also be achieved by very time-consuming processes, which have a low rate of material turnover, being an equivalent of sufficient material buffers on the input and output side.
Background
The ROI (return of investment) calculation is based on the productivity of a process. Beside the cycle time, it is also based on the available time in a given time frame. So one of the key factors to achieve the desired ROI is to look for the applications, which support maximum available working time for the robot implementation.
How to look at secondary savings?
Secondary savings are another important and so often underrated or overlooked factor, because some of them or their impact are not well understood.
Quality
One of the easier to understand secondary savings is the quality impact a stable robot process can achieve. The typical Monday morning quality is forgotten, instead an increase in overall stability as well as consistent production performance is realized. It simply takes human performance fluctuations out of the equation and therefore its impact on the process and quality.
Expenses
Some other factors seem obvious, but are also easily overlooked. Is your light in the production turned on? When a robot is running in the area alone, you can turn it off and save electricity costs. Or do you work in a controlled environment where access is controlled by air showers? The reduced number of entries based on reduced activity is another saving. To be frank, this is often easier to spot for an outsider with a fresh pair of eyes.
Who collects the Big Production Data?
In my opinion, the most overlooked feature of a robot is making the factory smarter by generating valuable data. In the simplest form counting parts going through the station, measuring cycle times, measuring material bottleneck times and so on.
Often there is the desire in the customer factory to achieve more insight and complicated systems are considered, which require a lot of additional hardware. But the point that the robot can be connected and used for needed data generation, is in only a few cases seriously considered.
This is a huge difference to a human-run station, which still tracks by hand in many cases in Thailand. Usually, it provides unreliable data, which does not allow correct conclusions at the same value. But probably more impacting the knowledge gained based on data of what happened with products, which show certain quality issues. The data allows tracking and evaluating or even resolving the yield loss at a station while the cobot is mainly deployed for a different task at the same time.
Do you face decreasing product life cycles?
Another more and more valued benefit of cobots is the flexibility of the asset compared to other machines or high payload robots. A lot of customers are dealing with increasing product changes or even product life cycles. So it is often not the best choice to use a single purpose machine as in many cases it can become obsolete after a year or two. This hits even harder, if the asset is under a BOI (Board of Investment) supported project, which requires you to use the asset for 5 or even 8 years.
There are a lot of areas where quick product changeover and especially fast roll out of the new product or process is what saves a lot of money and can turn into bigger sales. If you can manufacture faster and your customer is earlier to market, usually ahead of the competition, the overall sales during the product life cycle increase. This way not only do we select the robots for COGS (Cost of Goods Sold) reduction, but gaining additional profits as well.
Flexibility: Human vs. Cobot
Thailand is a hub for manufacturing, because labor costs are, or let’s say have been, low and the operator work quality is generally on a high level. A lot of production areas are labor-intense for the reason of flexibility, but this also has restrictions, like time to train operators for the roll out and higher yield loss for the ramp up period. Flexible cobot systems, in which one system gets tuned up to the desired quality and then rolling it out by copy and paste, are significant secondary savings which should be considered.
To even increase the benefits, the first cobot system should be in the RnD (Research and Development) or NPI (New Product Implementation) department, where the customer samples are made and the COGS are estimated. To be able to set up the actual production process at an early stage gives more insight for better decision making and an even faster rollout into production.
New Product Implementation Example
A good example for this is the manufacturing of latex products, like gloves for example. The actual production line is rather long for mass production, so new products are often made by hand. Using a cobot, which can repeat a process in the exact way over the number of samples and allowing to tune parameters while tracking the process parameters in the background, can make a huge difference.
The secondary saving here is not only the roll out savings, but also in many cases to not stop the main line for sampling of new products, causing an overall higher uptime for the main production line.
Recommendations & Best Practice
One recommendation is looking for processes where yield loss occurs frequently and can be linked to variations in operator performance. These are typically described as monotone repetitive tasks for which a cobot system is just more suitable.
Another recommendation is looking at things the cobot can do, which humans cannot do. Working in the dark or an environment which is unhealthy for humans or not suitable for a long period of time are good examples. Another example is starting early, ahead of the operator shift, based on a programmed start time or remote trigger. You can’t do it with humans without increased costs, which means there are savings to make.
The main recommendation however, is looking at significant benefits to the overall business from business intelligence by cobots collecting data as a secondary task. The data allows to improve quality response time and insights to increase yield, while giving the management more information for better decisions from new product COGS or production performance and capacities.
Background
We often see false ROI calculations, where a cobot investment is just compared against operator costs or even direct salaries only. It is important to look at all the possible savings and costs as well.
In many cases the secondary savings are driving the cost reduction over the whole investment and also open up new opportunities for the overall business in Thailand. Often especially when combined with Smart Factory initiatives or Continuous Improvement projects the cobots become a great first step or quick success because of things which couldn’t be done without them or the data they generate.
How to make the cobot implementation a success?
Our advice will differ depending on where you are on your cobot journey, whether you are new to cobots, have already started or have already finished an implementation and you are looking into getting more cobots or more benefits from your cobot implementation. There may be even some with unsuccessful applications, so it would be important to understand why it didn’t become a success the first time.
The most important part in general is to define the final goal. Often the goal is cost savings since this is an important driver for change in the way things are done. Our recommendation is to select the easier applications to get started and go for the bigger challenges at a later stage.
It’s important to keep the motivation and excitement that the cobots bring high and not letting them turn into frustration, because we took on a challenge we were not ready for yet. The fundamental principle to keep in mind is that if it doesn’t work out with the first try, we can always relocate the robot to a more beneficiary application, so we are not losing the investment and still make it a success.
The better way is to choose the best application and the best method of approach. What I mean is looking at available capacities and to decide whether it is the best to do the integration with internal resources or if it may be better to use external resources for the first or even all integrations.
For any of the described situations we will provide support in different ways. In general, our website will feature an increasing number of tutorials and how-to videos to support the integration up to full application step-by-step videos. Feel free to reach out on YouTube for example if you have specific requests or contact us via the contact page. In our videos we are often referring to recommended practices, which will make integrations a lot easier.
In case on site integration support is desired, we have a partner network in Thailand with distributors and system integrators, who can assist with the solution or provide a turn key setup. The partners are listed on the website as well with a short description, allowing you to find a match if you like to contact them directly or reach out to us for a recommendation.
Recommendations & Best Practice
We recommend having an open dialog with us directly or through our partners, so we can assist in evaluations of integration complexity, give recommendations on site regarding feasibility or provide the tutorials you actually need. In many cases, the best approach for cobot starters is to do at least the first integration together or go through the hands-on training with our standard application like CNC machine loading or PCB tester loading.
Background
With our vast experience of system integrations, support from AUBO China and our local Thai partners, we provide insights and best practices to make cobot applications even easier. Additionally, as a local business in Thailand we understand challenges our customers face. Having other cobot related products under the umbrella, we can provide coverage for all integration needs.
The Next Steps
Walk through your facility and keep an open eye and mind. Take notes and apply the guidelines given above. Feel free to contact us. Whether it is to consult you on your findings, to support you identifying the right projects (by providing training or as a service) or to realize your project for you. Use our contact form below to schedule a call in Thai, English or German language.
In case you feel like you want more information, browse through our News & Blog section to find the valuable information that you are looking for. If you are looking for specific information that is not there yet, use the contact form below to let us know, so we can work on your topic.