With the advent of Industry 4.0, IoT, cobots, machine learning, etc., many industry players seem to be following the course of going full throttle on automation. In the past decade we’ve seen transformations on the production floor, the labor market, and the global value chain. We’ve seen numerous companies embrace the move to “automate” and there’s no shortage of new manufacturers who would like to get started. When there is such a high demand, there are also endless vendors offering to help.
The race to automation in manufacturing hasn’t exactly been smooth-sailing for everyone however. There are too many examples of people charging full speed with regard to automation, both on the hardware and the software side, which sometimes end up causing more pain and hassle than what they’re actually worth. This is because they’re essentially eliminating the human factor and the inherent variability that can occur in the manufacturing operation for the sake of automation. In these cases, automation ends up being a restriction rather than a true improvement in productivity, quality and more importantly flexibility.
Let’s take two key categories of automation, one in the hardware side and another in the software side. Starting with the hardware side, the best and probably the most public example of this is what happened in Tesla. CEO Elon Musk spent a significant amount of funds to automate the assembly process in manufacturing the new Model 3 electric cars in 2018. Months after the process was launched, it became clear that it would take more than automation for Tesla to hit its manufacturing targets. Numerous changes and unforeseen scenarios in the manufacturing floor as well as in the products being produced meant that the machines and robots built specifically for each function and step in the production process would not work and the envisioned smooth process could not be executed. Musk himself admitted that “excessive automation at Tesla was a mistake” and that “humans are underrated.” Musk ended up scrapping what he called a “crazy, complex network of conveyor belts,” and drafted in human workers in an attempt to meet their deliveries.
Another example is at multiple plastics facilities across South East Asia that produces containers for the FMCG industry. Once the containers are made, they will be packed in boxes and shipped off to the fillers. Normally, the tedious process of packing the containers into boxes are done by hand. Naturally when Industry 4.0 sales folks come through, the recommendation for using a robotic arm to perform auto packing is offered. On face value, this is a fantastic idea. Automating this process would be so much more efficient, it would be faster, better safety, and workers would not have to handle freshly made hot plastic containers. With this in mind, a pilot is kicked off and the first line to automate would involve installing a robotic arm to pick-up and pack purple cylindrical 300-ml bottles. With the support of the vendor’s engineers this is proven to work great and after a showcase to the company directors, the vendor has done their job, and goes on their merry way.
A month or so later, product changeover begins. A new white rectangular 450-ml bottle is to begin production on the same line the robotic arm is installed in. A few potential events may now occur:
The manufacturer was wise and planned far enough ahead to have purchased an easily self-configurable solution to cater for a whole breath of sizes, colors and shapes and was fully tested with their own staff and validated prior to the vendor leaving.
Pray that the last training that the vendor did for the team stuck and that their own in-house team will be able to handle the change over and configure the system.
Call the vendor back and start reaching for that wallet to help with the transition.
Scramble to call Human Resources to try to hire someone familiar with the automation and vision system in place.
As the board of directors had already seen the showcase, push the robotic arm aside and bring in the original manpower to perform the manual packing.
Sadly case #1 doesn’t happen that often. What we end up with typically is a very costly and shiny new device sitting in a corner collecting dust as production must continue.
One key thing to note is that automation in both hardware and software is mandatory to stay competitive in this global environment. Despite the negative stories above, automation is a tool and depending on how you use it can either help or hurt. Engagements into automation requires a thorough analysis on flexibility, scalability and sustainability. All of which if not properly addressed, can lead to a lack of ROI, increasing cost of ownership over time, project failure or being stuck in pilot purgatory without any hope of being practically applicable throughout the facility and the enterprise. Therefore, the moral of the story is to not shy away from automation, but to remember to automate in moderation.