Automated inspection : Principles and methods

Group members: 

Ganesh Mokate 45

Gaurav Mane 42 

Rehan Sayyed 65 

Prathamesh Borude 8

Rohan Thorve 55

 INTRODUCTION

Testing has become an integral part of any production system. It is a form of rejection that does not go hand in hand with ensuring quality products. The advent of technically updated testing equipment helped to overcome the problems associated with traditional methods. The traditional approach has used labor-intensive methods that have led to an increase in lead time and production costs. In addition, there is a significant delay in obtaining an uncontrollable limit. Products that do not meet certain standards therefore incur additional costs for disposal and recycling.

The new approach to quality control sets the conditions for testing. The new route includes:

(i) Automatic testing method performed with 100% automatic testing.

(ii) Subsequent offline testing by installing online sensor systems for testing during or immediately after the production process.

(iii) Control of the production response function where a flexible process that determines product quality is more concerned than the product itself.

(iv) Mathematical process control is ensured using software tools to track and analyze sensor measurements over time.

(v) Advanced test and sensory technology, connected to a computer

programs designed to alter the functioning of the nervous system.

FOUNDATION FOUNDATIONS

Term testing can be defined as the task of evaluating products, parts, assemblies, or products, and determining whether they are compliant with construction specifications. Details of the design are set by the product designer.

Types of Testing

The classification of tests is based on the amount of information based on the test process on the compliance of the item and its specification.

Flexibility testing, in which case, appropriate measuring instruments or sensors are used to measure one or more quality markers.

Symbol testing, in which parts or products are tested to comply with the quality standard tested. Determination is sometimes based on an inspector's judgment. The attribute test includes calculating the number of errors in a product.

Typically, signal testing uses the P-chart and C-chart while the dynamic tests use the X-and R-chart.

Test Process

The steps in the test performed on each item, as part of, part or final product are as follows:

Item submitted for testing.

Testing

Item tested with incompatible features. Quantitative measurements or other features of a component or product are evaluated, while flexible analysis is performed.

The decision

Based on the test, a decision was made as to whether the item complies with the specified quality standards. A simple case involves a binary decision, in which case something is considered acceptable or unacceptable.

Action

Action should be taken in terms of the decision to accept or reject an item, or to set things up at the most appropriate level. Naturally, the advent of the automated testing system has put the test process in the back seat because of the benefits gained by industry with accuracy and time saving. In some production situations, the testing process is applied to only one object (e.g. a particular type of machine or type). In batch system and mass production, either bulk testing is done (so-called testing) or the sample is taken in bulk (sample testing). When testing one object or a few samples, it is done by hand and for mass production, automated systems are used for 100% testing. It is a time-consuming and expensive task to check all the sizes and features of the parts. Specific sizes and specifications are required for product blending or performance and are referred to as key markers (KCs).

 Inspection Accuracy

 Test accuracy

 There are usually two errors made by the manufacturers during the testing process. These two types of errors are called Type I and Type II errors. Type I error occurs when a lot is rejected and is called a producer risk. Type II error occurs when a lot of negative acceptance is accepted and is called consumer risk. There was an error by the inspector who missed an error while checking the assembly line.

In manual tests, these errors come from things like:

(i) Difficulties and difficulties encountered during the assessment process.

(ii) Biodiversity in the assessment process.

(iii) The need for adjudication by a human inspector.

(iv) Mental fatigue.

(v) Misconceptions or problems with gauges or measuring instruments used in the testing process.

After establishing the default system method, test errors occur due to features such as:

(i) The complexity and complexity of the assessment task.

(ii) Adjustment of sensitivity sensors affected by “profit” and set of common control parameters.

(iii) Equipment inefficiency.

(iv) Errors or "bugs" in the computer system that control the testing process.

The power of the testing process to avoid these types of errors is called testing accuracy. Test accuracy is high when few or no errors are made. Drury (1992) suggested steps for the accuracy of a case test where the parts were divided by the examiner (or the default test system) into two categories, parallel or inconsistent. When you think about this binary case, let p1 be a chance for you to make the right decisions. Therefore (1p1) it is possible that the corresponding object is classified as inconsistent (Type I error). Similarly, if p2 has the potential for an invariant object to be classified as inconsistent at the time, (1p2) it is likely that the inconsistent object is classified as sync (Type II error). things. A table of potential outcomes can be created as shown in the Table to indicate the fraction of the components that are positively and negatively separated and those that are incorrectly classified, even if the error is Type I and Type II.

Ideally, these estimates (probability) will be evaluated individually by determining the appropriate number of decisions made in each of the two cases of compliance and non-compliance in certain interest categories. Unfortunately, sizes vary with different test functions. The most difficult test tasks usually have high error rates (p1 and p2 values ​​are low). Also, p1 and p2 values ​​are different for different testers. Standard values ​​for p1 grades are between 0.9 and 0.99 and those for p2 between 0.80 and 0.90, but values ​​as low as 0.50 for both p1 and p2 have also been reported (Drury 1992). p1 tends to be higher than p2 for human testers because inspectors tend to check for errors and check for good quality

Inspection vs. Testing

Quality Control (QC) utilizes both inspection and evaluation processes that are equally important in the company's quality control system. Testing is a term for quality control that refers to the evaluation of the functional properties of a product and testing is used to assess quality and specificity of its design. An item tested in QC testing is recognized during real-time operation or under possible conditions during operation. For example, to know if a product is working well, you are tested, by using it for a certain period of time. In some cases, the testing process is a waste of time, in which a number of items are determined to ensure the quality of the material. Efforts are being made to develop methods known as non-destructive testing (NDT) and non-destructive testing (NDE) to save costs incurred during invasive testing.

SAQ 1

(a) Discuss the various steps involved in the assessment process.

(b) What does he mean by the accuracy of the test?

(c) What are the errors of type I and type II?

AUTOMATIC TEST

Test Program

In the current scenario, manual manipulation is significantly replaced by automatic testing as errors are significantly reduced by the process automatically. The economic adjustment of the automated assessment system depends on the cost savings of staff and the improvement in accuracy will be greater than the cost of investment and / or development of the system.

Automatic testing is defined as the automation of one or more steps involved in the testing process. Automatic or automatic testing can be done with a number of alternatives.

(d) Automatic introduction of components through an automated management system with manual testing and decision-making processes.

(e) A machine with automatic loading parts for the machine that performs, automatic testing and decision making.

(f) Complete automatic testing system for the presentation of components,

tests and decisions are performed automatically. The testing process is done by the first person working, and all possible errors can be made with this test method. In the second and third cases, the actual test performance is achieved by default. Like manual testing, automated testing can be done using mathematical examples or 100% tests. Sample errors can occur when using mathematical models. Like a human tester, an automated system can make a sample test or a 100% test. Human testers can make such mistakes. The default system works with the highest accuracy of simple test tasks such as default

simple size estimation for a particular component. With the increase in the complexity of testing, the error rate often increases. Some machine vision applications fall into this category; for example, finding faults on integrated circuit chips or printed circuit boards. PCB testing tasks are complex and difficult for human staff. This is one of the reasons why building automated testing programs can perform such tasks. As mentioned earlier, test errors can be classified as Type I and Type II.

Type I error occurs when the default system displays a false alarm while the process is being controlled.

Sensitivity of other automated testing systems can be adjusted by adjusting their “profit” or similar controls. When sensitivity correction is low, the chances of type I error are low but the chances of type II error are high. As the sensitivity correction increases, the chances of type I error increase, and the chances of type II error decrease. This relationship is described in Figure. Due to these errors, the test system cannot guarantee a 100% quality product.

The full potential of automatic testing is best achieved when combined with a production process and using 100% testing, even when the results of the process lead to some positive action. Good deeds can take either way possible, as shown in Figure

Feedback Process Management

In experimental operations, data is a response to the production process that is responsible for that and the same test is performed. The motive for the response is to allow for compensation in the process of reducing diversity and improving quality. If the process effect starts to slide toward a higher tolerance angle (e.g. wearing a tool can cause the size of the part to increase over time), adjustments can be made to the input parameters to return the result to a higher value. In this way, the average quality is maintained within the range of less variation than is possible by means of testing the examples. As a result, process power is improved.

Division of Parts

Here, the components are classified on a quality basis and categorized as an accepted or rejected component. There may be more than two standards of process-appropriate quality (e.g. acceptable, renewable, and waste). Filtering and testing can be done in many ways. One way is for both of them to check and edit on the same channel. Some inputs receive one or more tests in the processing line, with a single filter channel near the end of the line. The test details are checked and the test is sent to a screening station indicating which action is required for each component.

Test Time

An important consideration in quality control is the timing of the evaluation process. Three different options can be seen as shown in the Diagram

namely:

(a) illegal inspection,

(b) online / program, and

(c) online / testing process.

Offline Testing Methods

In offline testing, test equipment is usually provided and does not touch anything with the machine tools. There is always a time delay between production and testing. Manual tests are commonly used to encourage the use of offline testing which includes:

(i) the process variability is within the design tolerance,

(ii) processing conditions are stable and the risk of significant deviation from the process is small, too

(iii) the costs incurred during the inspection are higher compared to the costs of a few defective components.

The disadvantage of offline testing is that the components have already been performed at the wrong quality. Sometimes by default the defective part may not be included in the sample. A correlation measurement machine (CMM) is an example of offline testing. CMM is discussed in more detail in the next section.

On-line / In-process and On-line / Post-process Test Methods If the test function is done as parts are done, then it is called the Internet. There are two variations of online testing. If the test is performed during production, it is called on / line-in-process inspection. If the test is performed immediately following the production process, it is called an on-line / post-process test as shown in Figure in the machine tool spindle, or stored in a tool tool that can be exchanged with automatic tool just as the tools are handled. In flexible production machine tools the focus spinning tools are widely used. The main test components of sensory investigations. Symbols are transferred to the controller as the contact is created

half part. More technology is available to transfer signals. Some of them are direct electrical connections, induction coil, infrared data transmission. The task of data processing and translation is simplified by the controller.

SAQ 3

(a) What does it mean by automatic testing?

(b) Write down the steps involved in automating the industrial inspection process.

(c) What is the management of the accountability process?

(d) Distinguish between on-line / in-process and on / line / post-process testing methods.

The Three Types of Quality Inspections

There are three primary types of quality inspections: pre-production, in-line, and final. There are a variety of details that must be inspected and approved during each phase in order to detect and correct quality problems.

Pre-production Inspection

During the pre-production phase, raw materials should be tested before entering production. This may include a number of tests to examine the material for weight, dimensional stability, pilling resistance, torqueing, pile retention, stretch recovery, and much more. Components including closures, zippers, elastics and other embellishments such as beads, rhinestones, sequins and rivets should also be tested for regulatory requirements.   

Since quality issues are often a result of defects in the materials, inspections during the pre-production phase allow auditors to address any issues before production begins. Ultimately, by inspecting the materials up front, brands and retailers can avoid unanticipated costs and delays.

In-line Inspection

Additional inspections should take place during various stages of production. For apparel, inspections should occur at each critical step of the production process, from cutting to assembling to pressing or other finishing procedures. For example, during the cutting phase, each cutting ticket should be randomly inspected to ensure that each part is accurately notched and shades are separated. If fabric is incorrectly cut, the parts cannot be properly assembled.

In-line inspections are important, as quality issues are often re-workable during the production phase and can be fixed before the final product is complete. When quality issues are not corrected during the production process, minor issues in the beginning of production can lead to larger issues in later stages.

Final Inspection

The final inspection is the last opportunity for auditors to catch and address quality issues before they end up in the hands of the buyer, or even worse, the consumer. During the final audit, products are examined for specific performance requirements, overall appearance, sizing and fit.

Brands and retailers often skip inspections while the products are still at the factory and only perform random, final inspections once the order is received at the ultimate destination. By then, it is too late and the only recourse is discarding the poor quality units. This is costly to every party, especially the factory, which will bear the brunt of the expense.

 CONNECT MEASUREING MACHINERY

The measurement of the original nature and size of an object and its comparison with the required shape and size as described in the drawing section enters the broad area of ​​Coordinate Metrology. Exploration of space, shape, size and geometry and part or objects are various aspects of metrology. An integration measuring device is an electromechanical system designed to perform the metrology of a connection. CMM contains a communication probe, and this communication probe is placed in three (3-D) areas related to the surface of the work part. To find out, the largest data pertaining to the geometric component, x, y, z probe links are accurately measured.In a three-dimensional link system, the basic CMM is made up of the following components.

(i) Investigate the topic

(ii) Mechanical and transducer movement,

(iii) Drive system and control units, and

(iv) A computer program containing application software

This unit oversees information on the basics of testing and their relationship to the level of customer service set. There is enough focus on the various types of testing process, and accuracy, especially to highlight the importance of automatic testing, which has the potential to reduce lead time and cost. The integration of the measurement machine and related techniques is discussed in this unit to show the growing area of ​​use of computers with contact testing techniques

Vision guided robotic inspection system technology 

Recently, 3D vision guided robotic inspection systems became industrial reality, due to the advancement of their control systems incorporating powerful parallel processor-enabled PCs with large internal caches, and high performance external memory interfaces for balancing the example Braintech has developed a unique single conventional CCD video camera-based 3D vision technology for robust 3D robot guidance and inspection for automotive assembly and inspection applications (Figures 4 and 5). (This technology is intended mainly for rigid parts with stable dimensions and is not suitable for flexible parts whose dimensions vary significantly such as plastic blow-molded products.) The key with any vision guided robotic inspection technology is to seamlessly integrate the 3D vision side with the robotic control side. Some highlights of this integration challenge include the following

The clear dialog boxes and menus in the user interface programs allow the engineers to closely tailor the communication parameters to the needs of each application (Figure 4). The result is reliability, speed, flexibility and simplicity (Scott, 2002). The architecture contains a host of 3D calibration, 3D geometry, object location and robot coordinate system transformation components specifically designed for robot guidance. This tight link to robot operation results in high accuracy and repeatability; all vital to automated, robotic assembly, guidance and inspection. Figure 4 An example of the dialog boxes and menus in the user interface program of the Braintech vision guided robotic system Assembly Automation Volume 23 · Number 3 · 2003 · 252–256 Figure 5 The Braintech vision guided robotic part handling and inspection system, based on a single, conventional CCD video camera integrating 3D vision technology for rigid parts with stable dimensions .

 All robot and PLC communication is handled through specialized Read, Write, and Poll components. The complicated details of the underlying protocols become transparent. Alternative solutions to the single conventional CCD video camera-based 3D vision include stereovision, and laser triangulation methods. The goal of the stereovision technology is to calculate the depth or distance of features or landmarks on a given object relative to the sensor (i.e. to construct a depth map). The depth information from features or landmarks can then be used to calculate the 3D pose of a given object. This technology uses images from dual cameras aimed at the same object. Stereovision algorithms operate by locating the same features in both images. Using the geometrical relationship between the two cameras and the location of the feature in each image, the depth of each feature can then be triangulated and a depth map can  be constructed. Ideal applications for stereovision involve cases when the given object is not rigid and its dimensions are not Another method, the laser triangulation 3Dtechnology, uses the same principle as the laser targeting systems use in aerial combat by military aircraft to target air or ground objects. By painting a surface using a laser beam (structured light), the laser triangulation sensor determines the depth and in some cases, the orientation of the surface being observed. The sensor is typically an integrated, hardened casing that houses a CCDcamera and a low-power laser diode. More sophisticated versions of this technology feature scanning lasers that project a plane (instead of a spot) onto the surface of the object. The laser plane projection and its degree and direction of distortion can then be analyzed to render orientation information about the surface in question.

Summary and conclusions

Due to the advancement of 3D vision systems and related processing power, we are not far from having “invisible operators” working along assembly lines, inspecting every process in a real-time, feedback-controlled environment for the purpose of keeping the processes within their statistically pre-programmed limits. The key though, as always, is not to miss alternative inspection technologies, that could do the job well at the fraction of the complexity and therefore the cost, and thinking even further, by designing quality into the product, to reduce the need for inspection.