Fixtureless Robotic Assembly Using AI Virtual Datums

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

The high costs and obsolescence of dedicated hardware fixtures in manufacturing, particularly in the automobile industry, due to the complexity of design and frequent product changes, necessitate a more flexible and cost-effective solution for securing and locating sheet metal parts during welding processes.

Innovation Solution

A reconfigurable, fixtureless manufacturing system utilizing learning A.I. software and material handling robots with gripping elements, which can adapt to various part shapes and sizes, eliminating the need for physical fixtures by using machine vision for alignment and spatial adjustments, and storing virtual datums for optimization across multiple production iterations.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Manufacturing precision

If dedicated hardware fixtures are used to secure and locate sheet metal parts, then manufacturing precision and stability are improved, but device complexity and cost increase significantly

Engineering Contradiction:
Improvepart positioning precisionVSAvoidfixture design complexity
Core Design Contradiction:
Manufacturing precisionVSDevice complexity

Solution Approach 1:

The patent replaces the mechanical fixture system with a robotic manipulation system equipped with sensors and control algorithms. Robots with specialized end-effectors grasp and position parts based on visual feedback and force control, eliminating the need for complex mechanical fixtures while maintaining positioning precision through active control rather than passive mechanical constraints.

Inventive Principle:
Principle #28Mechanics substitution (Replace mechanical system)

Solution Approach 2:

The patent uses digital models and virtual fixtures that replicate the positioning function of physical fixtures. Through computer vision and haptic feedback systems, the system creates virtual representations of part geometries and assembly constraints, allowing robots to locate and position parts accurately without physical datum features or fixtures.

Inventive Principle:
Principle #26Copying

2Manufacturing precision

If dedicated hardware fixtures are designed for each subassembly, then manufacturing precision is improved, but adaptability to product changes deteriorates

Engineering Contradiction:
Improveassembly alignment precisionVSAvoidproduct change adaptability
Core Design Contradiction:
Manufacturing precisionVSAdaptability or versatility

Solution Approach 1:

The patent employs dynamically adjustable robotic systems that can adapt their grasping and positioning strategies in real-time. The robotic manipulators can modify their end-effector configurations, grasp forces, and trajectory plans based on detected part variations, enabling the same system to maintain precision across different product configurations without retooling.

Inventive Principle:
Principle #15Dynamics

Solution Approach 2:

The patent implements a universal robotic assembly system that can handle multiple part families and assembly configurations through programmable control. The system uses vision-guided recognition, adaptive grasping algorithms, and reconfigurable end-effectors to perform the same positioning and assembly functions across different products, replacing the need for dedicated fixtures for each subassembly.

Inventive Principle:
Principle #6Universality (Multi-functionality)

3Adaptability or versatility

If flexible fixture systems are used to accommodate product changes, then adaptability is improved, but device complexity and cost increase

Engineering Contradiction:
Improvefixture reconfiguration capabilityVSAvoidfixture system complexity
Core Design Contradiction:
Adaptability or versatilityVSDevice complexity

Solution Approach 1:

The patent replaces the mechanical reconfiguration mechanisms of flexible fixtures with software-based control of robotic systems. Instead of physically adjusting fixture components to accommodate product changes, the system uses programmable robot motions, vision system recalibration, and control algorithm modifications to adapt to new parts, eliminating mechanical complexity while maintaining reconfigurability.

Inventive Principle:
Principle #28Mechanics substitution (Replace mechanical system)

Solution Approach 2:

The patent achieves adaptability by changing control parameters and software configurations rather than physical fixture parameters. The robotic system adjusts grasping forces, positioning tolerances, motion trajectories, and vision system calibration parameters through software to accommodate different products, replacing the need for mechanical fixture reconfiguration.

Inventive Principle:
Principle #35Parameter changes

4Productivity

If robots and automated assembly systems with physical fixtures are used, then productivity is improved, but loss of time for fixture updates increases

Engineering Contradiction:
Improveassembly automation levelVSAvoidfixture update time
Core Design Contradiction:
ProductivityVSLoss of time

Solution Approach 1:

The patent replaces physical fixtures with robotic manipulation systems that require only software updates for product changes. The robotic system maintains automated assembly capabilities while eliminating the time-consuming process of physical fixture installation, adjustment, and validation, as adaptation is achieved through reprogramming rather than retooling.

Inventive Principle:
Principle #28Mechanics substitution (Replace mechanical system)

Solution Approach 2:

The patent implements digital twins and virtual commissioning processes that allow fixture configurations to be developed, tested, and validated in virtual environments before physical implementation. This preliminary digital preparation reduces the time required for actual fixture updates or robot reconfiguration by pre-resolving potential issues and optimizing parameters beforehand.

Inventive Principle:
Principle #10Preliminary action

Data Source

PatentUS11559897B2Reconfigurable, fixtureless manufacturing system and method assisted by learning software
Publication Date: 2023.01.24 GHANEM GEORGE K
  • US11559897B2 patent drawing
  • US11559897B2 patent drawing
  • US11559897B2 patent drawing

AI summary

Systems and methods for AI assisted reconfigurable, fixtureless manufacturing is disclosed. The invention eliminates geometry-setting tools (hard points, pins and nets—traditionally known as 3-2-1 fixturing schemes) and to replace the physical geometry setting with virtual datums driven by learning AI algorithms. A first type of part and a second type of part may be located by a machine vision system and moved by material handling devices and robots to locations within an assembly area. The parts may be aligned with one another and the alignment may be checked by the machine vision system which is configured to locate datums, in the form of features, of the parts and compare such datums to stored virtual datums. The parts may be joined while being held by the material handling devices or robots to form a subassembly in a fixtureless fashion. The material handling devices are able to grasp a number of different types of parts so that a number of different types of subassemblies are capable of being assembled. The system enables one skilled in the art to develop a product design with self-locating parts that will eliminate and minimize the need for geometry setting dedicated line tools and fixtures. This leads to the development of a manufacturing process that utilizes the industry 4.0 technologies to once again eliminate or significantly reduces the need for geometry setting line tools.