A cement company stopped sending its personnel to inspect dangerous areas. Your new inspector is a robot dog
Even though the conversation revolves around humanoid robotics Lately, this sector has had much more traction in the industry for decades, driving a good part of the current processes and assembly lines. In this regard, a Swiss cement plant has wanted to take advantage of robotics in a somewhat peculiar way: it has been using a quadruped robot to monitor your facilities every night. Below these lines we tell you all the details. The problem that had to be solved. Vigier Ciment has been producing cement in the hills of the Swiss Jura, south of Biel, for a century and a half, generating approximately a fifth of all cement in the country. Its plant houses more than 1,000 machines spread across several buildings and floors, connected by metal stairs of up to 16 sections, areas with temperatures that reach 50 degrees, constant dust and the occasional presence of ammonia near the unloading docks. The maintenance of all this fell to operators who toured the facilities on foot filling out paper records. Over time, continued exposure to these conditions generates what the plant workers themselves call “operational blindness.” You stop seeing what is in front of you because you have already seen it too many times, according to collect Techeblog. The guard dog. Just like account The medium, in November 2024 the Swiss robotics company ANYbotics began talks with Vigier Ciment to deploy its ANYmal quadruped robot at the plant. The robot arrived on January 6, 2025 and before the end of the first month it was already carrying out night patrols completely autonomously. ANYmal is similar in size to a large dog and weighs more than 50 kilos. It does not need human supervision, and its managers say it climbs stairs, avoids obstacles, navigates narrow hallways and accesses areas that previously required considerable effort on the part of staff. What exactly does it do in every round. Every night, even on weekends, ANYmal goes through more than 450 inspection points predefined elements distributed in three mills and six levels. To do this, it has several detection systems, including a visual camera that identifies cracks, oil leaks or corrosion; a thermal camera that measures the temperature of critical components such as bearings, motors and gears; a gas sensor that monitors ammonia levels; and an acoustic camera capable of locating compressed air leaks or filter failures at distances of up to 50 meters. According to point In the middle, all that information is automatically dumped into a software platform called Data Navigator, which analyzes the data collected overnight, compares it to the facility’s history, and generates a daily report for the maintenance team. What he has found along the way. In sixteen months of operation, ANYmal has already completed more than 33,000 inspections without recording any technical failures. According to ANYbotics, the most relevant findings have had a direct impact on plant operations. The middle share In addition, the robot detected a crack in the base of a shredder the size of a large kitchen table. The oil had been leaking for some time and no one had reported it on the usual rounds. The repair was completed the next day. Had it collapsed, the plant would have lost more than a week of production, with estimated losses of more than $630,000, according to the company’s own figures. In another case, thermal monitoring detected a bearing reaching 140 degrees Celsius, allowing a $30,000 eight-hour repair to be scheduled rather than facing a much more costly emergency failure. The robot also detected levels of ammonia exposure at unloading docks that had not been measured until then, and located air leaks in filtration systems installed 50 meters high. Industrial maintenance. The plant’s traditional fixed sensors only covered about 200 elements, mainly on the clinker side (the main component of cement). The robot expands that coverage substantially and accesses areas that static sensors cannot reach. At the same time, it removes operators from the most dangerous environments without reducing the frequency or quality of inspections. Images | ANYbotics In Xataka | One of the big problems with AI is that it always proves you right: this is the most effective way to avoid it
