This article was published in Scientific American’s former blog network and reflects the views of the author, not necessarily those of Scientific American
Most people think about robots as autonomous machines guided by artificial intelligence. Most people are wrong: robots come in all sorts of forms, which I know because I am a roboticist in a cheese factory. I work with robots fully dedicated to cutting, shredding, sorting and boxing up cheese.
It wasn’t by accident that I ended up working on robots in a cheese factory. In fact, most of my education has been leading up to this career although I didn’t know it at the time. I completed my BA in linguistics (from University of California, Santa Cruz) in 2013 with a primary focus on the theoretical side of syntax and semantics. I spent most of my time parsing the English language into its fundamental components.
But then I quickly turned from the liberal arts to the industrial arts by enrolling in machining school. I learned to run manual metal-cutting machines and program the computer-controlled versions. The syntax of G-code (a programming language ubiquitous to machining) is transparent compared to the syntax of natural language. My training in linguistics prepared me well for this.
On supporting science journalism
If you're enjoying this article, consider supporting our award-winning journalism by subscribing. By purchasing a subscription you are helping to ensure the future of impactful stories about the discoveries and ideas shaping our world today.
While gainfully employed in a high-production machine shop, I saw a real need for automation in manufacturing. After the shop where I was working sustained a layoff prompted by a lull in the semiconductor industry, I jumped at the opportunity to pursue a career in automation. I began taking mechatronics courses in January 2017. It was only three months into the program when I was approached by my current employer, Lactalis, to work on automated equipment at a cheese factory. That is how I found myself working at one of the largest cheese factories in the world.
Robots in a cheese factory do not think on the fly, so to speak. They do not decide what to do next. Each robot is programmed to perform the same task over and over again.
For example, we employ a palletizer in the whey plant. Milk is separated into solid curds and liquid whey while making cheese; the curds are used to make cheese whereas the remaining liquid is dehydrated into powder. A palletizing robot uses an arm to pick 50-pound sacks of whey powder off a conveyor, neatly arranging the sacks onto a pallet. Once enough sacks have been stacked, the arm picks up a roll of industrial plastic wrap and secures the stack so that the pallet remains neatly packed and undisturbed during shipping. The number of sacks and height of stacking is called out in the palletizer’s programming.
A cheese factory needs robots to perform repetitive tasks at high volume 24 hours a day. The most common robots found in the cheese factory are large industrial machines dedicated to specific processes in the cheese-making line. These robots run autonomously and can be more than six feet tall and are sometimes 30 feet long.
Many of these machines are equipped with sensors to “see” the cheese. This is useful because if the cheese is not yet in position, it would not make sense to do the next step in the process. Once a sensor is flagged, signaling that there is cheese, the machine executes the next step until all stages of the process are complete.
For one process in particular, packaging string cheese, the machine is designed to seal 16 strings of cheese automatically without any intervention.
However, the loading process is not always perfect. Sometimes strings of cheese are missing. Although the machine can seal 10 strings as easily as 16, there will be six empty plastic packages after the sealing process—which cuts down on output and efficiency. But the machine hardware is fixed, and always seals 16 packages at a time, full or not.
A counting sensor is an elegant solution to this kind of problem. As string cheese moves into the sealing station, a sensor is keeping track of the cheese. If the count does not reach 16, a buzzer sounds, signaling that there is missing cheese, and sealing does not execute. Operators monitor each machine’s function. In the above scenario, an operator is responsible for replacing missing cheese and uses an override switch to resume the sealing process.
It is my responsibility as a roboticist at the cheese factory to maintain the automated equipment and ensure equipment efficiency. The empty package sensor, for example, is an aftermarket modification put in place by a technician like myself; the equipment originally did not come with any cheese-counting sensors. To integrate the sensor into the sealing process, the technician reprograms the robots accordingly.
It is important to note that automation is only useful if the outcome of the process is more efficient than it would be with humans doing the job. For that reason, it is one of my duties to improve and update our existing automated equipment.
In fact, I will complete robotic arm certification this year. And unlike the whey plant, the string cheese department—the department in which I directly work—does not yet have a palletizing robotic arm.
That’s something I hope to soon change.