In the distribution centers from which the millions of products that we buy in the shop and online start, to allow the positioning of the pallets at dizzying heights and the load of the trucks, the applications of Computer Vision think like those of an Italian, RoboSense start -up.
In modern warehouses, where millions of products are moved every day, precision is everything. The pallets – standardized wooden or plastic platforms with “pockets” for the insertion of the forks – are stacked in shelves up to 17 meters high (like a five -story building) and moved by autonomous elevators who work without operator. A few centimeters error is enough to drop a heavy load, with disastrous consequences for safety and costs.
The problem is more complex than it seems. Even a minimal inclination of the floor – less than one degree – can create a lever effect that is amplified throughout the height of the elevator cart, causing a deviation of 20 centimeters at the top. For autonomous elevators, increasingly widespread in logistical centers, this millimeter precision represents an important technological challenge.
The Italian solution
To remedy all RoboSense, technological reality based at the Business Innovation Center in Trentino Development (Pergine Valsugana, TN), has developed Smartfinder, an advanced system that makes use of an industrial camera located near the forks and sophisticated artificial vision algorithms (the so -called “computer vision”), which allow the cart to be clearly identified of insertion in the pockets of the pallet. The system is able to perform real -time corrections to compensate misalignments due to the irregularities of the floor, as well as to automatically recognize the size and model of the pallet between the standard ones of the EU and those in use in the UK and USA.
The precision of the system lies in the owner software developed by Robosense, which uses the potential of 3D Tof (Time-Of-Flight) cameras, which calculate the distance between the optical sensor and the objects framed in real time, allowing a three-dimensional mapping of the environment. “This technology allows a qualitative leap in the perceptive capabilities of mobile robots, bringing automation and operational intelligence to unpublished levels so far”, comments prof. Mariolino De Cecco, professor of industrial engineering and director of the Miro Lab Laboratory (mechanical and thermal measures) of the University of Trento, among the founders of the first startup from which Robosense originated.
However, automation in the logistical sphere stops in the Smartfinder system. RoboSense has recently created two other solutions that aim to make logistics centers more efficient and more and more automated: Smartgate integrates three 3D Tof cameras and a software module to analyze the pallets in transit on transporters, and is able to estimate the volume of the load and detect any protrusions or misalignments compared to the base, which could indicate anomalies in the packaging or potential handling, thus requesting a manual verification; Atl (Automated Truck Loading) is instead a solution for the loading of transport vehicles: based on Lidar technology, which uses laser impulses to determine the distance from the objects and build three -dimensional models of the surrounding environment, it is able to identify the exact position of the internal walls of the truck, their orientation compared to the load bay and the presence of any obstacles.
The system thus allows you to optimize the loading phase, reducing the risk of collisions, improving operational efficiency and increasing the overall level of safety. So that the goods arrive intact and quickly to the destination.
