Learning software assisted fixtureless object pickup and placement system and method

A method for learning software assisted, at least partially fixtureless assembly of parts into an assembly, said method comprising:

• deriving, by way of a controller, measured target points from one or more images of a part in a work area received from a machine vision system positioned to view the work area;
• performing, by way of the controller, an iterative analysis to determine a best fit solution for moving said part adjacent to another part for forming the assembly such that each of said measured target points is within a predetermined range of an associated one of desired target points for said part, where the other part is located at a support object or objects within the work area;
• converting, by way of the controller, said measured target points to a common reference frame;
• and
• moving, by way of electronic command issued from the controller, said part, by way of the robot, according to said best fit solution as converted such that said part is moved and secured adjacent to the other part in a fixtureless manner.

A system for learning software assisted, at least partially fixtureless assembly of parts into an assembly, said system comprising:

• a robot;
• a machine vision system positioned to view a work area;
• one or more support objects within the work area;
• a controller in electronic communication with said robot and said machine vision system;
• and
• one or more databases provided at, on in electronic communication with, said controller, wherein said one or more databases comprise desired target points, at least some of which are associated with weightings;
• said controller comprising software instructions stored on one or more non-transitory electronic storage devices which when executed, configure one or more processors to:
• perform, by way of the machine vision system, an initial scan of the parts to identify a candidate part of a specific type, which are stored at a bin within the work area, wherein said bin contains different types of parts;
• derive measured target points from one or more images of the candidate part in the work area received from the machine vision system;
• perform an iterative analysis to determine a best fit solution for moving said candidate part adjacent to another part of different type for forming the assembly by using a multi-factor vector fitting technique and at least one artificial intelligence technique such that each of said measured target points is within a predetermined range of an associated one of desired target points for said candidate part, where the other part is located at said one or more support objects, and where, based on one or more images received from said machine vision system, said target points of said candidate part cannot be placed within said predetermined range of said desired target points, adjust said weightings until said target points of said candidate part are placed within said predetermined range of said desired target points;
• perform an iterative analysis to determine a best fit solution for moving said candidate part adjacent to another part for forming the assembly such that each of said measured target points is within a predetermined range of an associated one of desired target points for said candidate part, where the other part is located at said one or more support objects;
• convert said measured target points to a common reference frame, where said common reference frame designates a location at said one or more support objects as an origin;
• command movement of said part, by way of the robot, according to said best fit solution as converted such that said candidate part is moved and secured adjacent to the other part in a fixtureless manner such that said candidate part is supported exclusively by the other part and/or the one or more support objects;
• and
• command joining of said candidate part and the other part, by way of the robot or another robot, while the candidate part is supported exclusively by the other part and/or the one or more support objects, at least temporarily.