Tweet
Background
When an offer is made by a SolidCAM customer for a CNC Job, the Shop manager makes an estimate of the material, tool costs, machining time, either based on experience (easier for 2D operations, much more difficult for 3D operations) or doing some of the operations on SolidCAM to get simulation time, and he gives an offer based on that.
If he gets the job, then he does all the CAM programming using SolidCAM and generates the CNC operations and the approximate simulation time. After setup, he runs the CNC operations to get the real-time through manual measurement. Then he could plan his CNC production, based on these actual times monitored during the first part run.
Application
The SolidShop application, based on Computer Vision, gets from one side as input the SolidCAM-defined CNC operations and their simulation time, and the first Part actual machining time. We use computer vision to monitor the actual CNC machine cycle (by detecting tool changes) for further Parts of the series, always comparing it with the First part machining cycle, thus locating abnormalities. Decisions can then be taken in the flow of the operations, based on the type of abnormalities.
Machine learning is used to analyze the time differences, collected for a period of time, between the first-part cycle and further cycles, for each CNC machine in the shop. This analysis should enable the user later to forecast accurately the actual machining time from the first-part time, for a specific material, CNC machine type, and CNC part.
SolidShop is linked to the CNC machine shop ERP and SolidFlow systems - SolidCAM Tool table data are also linked.
When an offer is made by a SolidCAM customer for a CNC Job, the Shop manager makes an estimate of the material, tool costs, machining time, either based on experience (easier for 2D operations, much more difficult for 3D operations) or doing some of the operations on SolidCAM to get simulation time, and he gives an offer based on that.
If he gets the job, then he does all the CAM programming using SolidCAM and generates the CNC operations and the approximate simulation time. After setup, he runs the CNC operations to get the real-time through manual measurement. Then he could plan his CNC production, based on these actual times monitored during the first part run.
Application
The SolidShop application, based on Computer Vision, gets from one side as input the SolidCAM-defined CNC operations and their simulation time, and the first Part actual machining time. We use computer vision to monitor the actual CNC machine cycle (by detecting tool changes) for further Parts of the series, always comparing it with the First part machining cycle, thus locating abnormalities. Decisions can then be taken in the flow of the operations, based on the type of abnormalities.
Machine learning is used to analyze the time differences, collected for a period of time, between the first-part cycle and further cycles, for each CNC machine in the shop. This analysis should enable the user later to forecast accurately the actual machining time from the first-part time, for a specific material, CNC machine type, and CNC part.
SolidShop is linked to the CNC machine shop ERP and SolidFlow systems - SolidCAM Tool table data are also linked.
Comment