Error Proofing Assembly Processes

Page 1:Error Proofing Assembly ProcessesManufacturing with ongoing quality and productivity✓✗
Page

Page 2:3Error ProofingAssembly ProcessesCommitment to ongoing quality total quality managementError Proofing: Key to QualityQuality demands on manufacturers have been high for years, andthe future will see those demands continue to rise. It is no longergood enough just to produce a high quality product. Expectationshave climbed to a point where that high quality product also has tobe delivered Just-in-Time (JIT), in the proper sequence (mainly forsub components), and provide a level of traceability that enablesvisibility to all stakeholders.Error Proofing is:TraceabilityQuality Just-in-TimeSequencingBearingBearing GreaseSpring WasherThere is no mistake – quality is an ongoing, fundamental part ofany manufacturing process. It means continuous improvement. Italso means that quality is everyone’s responsibility, it’s not just onedepartment or only done at the end of the process. Error proofingdevices, referred to as Poka-Yoke devices, such as sensors, need tobe integrated into the process with complete traceability throughout.Also, provisions have to be in place to quickly and easily add newPoka-Yoke devices and additional traceability as requirementsevolve.Error Proofing Assembly ProcessesThe Error Proofing Process4Step no. 1Identify Trouble Spots 6Step no. 2Implement a Detection Method 8Sensor Based Detection 10The Perfect Error Proofing Tool 12Color Sequencing 14Traceability for Just-in-Sequence 15Step no. 3Contain Product Discrepancies 16Add Complete Visibility to your Assembly Systems 18Supporting Systems Architecture 20ProductsIndustrial Networking and ConnectivityEconomical connection of sensors/actuators to the controller 22Industrial IdentificationBIS Industrial RFID Systems 24BVS Vision Sensors 26Object DetectionBFS True-Color-Sensors 27Photoelectric sensors with discrete and analog outputs 28Linear Position Sensing and MeasurementThe appropriate measuring principle for the optimal solution:optical, magneto-inductive and inductive 30Systems and ServiceCustomized Software for Handhelds 32Individually Programmed RFID Tags 33Customized to your specifications 34www.balluff.com

Page 3:4A Simple 1-2-3 Process:Identify, Detect, and ContainThe error proofing processDuring the development of the manufacturing line, trouble spotsare projected based on experience. However, during the life of amanufacturing line, additional trouble spots will inevitably surface.Simply identify those spots, implement a detection method, anddevelop a means to contain it.ContainDiscrepancies ImplementDeterctionIdentify Trouble SpotsTotalQualityManagement

Page 4:Step No. 1:Identify trouble spotsErrors in manufacturing and assemblyprocesses happen and need to be identifiedor predicted before you can take action.Step No. 2:Implement detectionActive error-proofing uses sensorsand vision systems to actively verify thata process is completed correctly.Step No. 3:Contain discrepanciesThere are three main ways to containdiscrepancies: scrap, reclassify,or rework the part.ContainDiscrepancies ImplementDeterctionIdentify Trouble SpotsTotalQualityManagementContainDiscrepancies ImplementDeterctionIdentify Trouble SpotsTotalQualityManagementContainDiscrepancies ImplementDeterctionIdentify Trouble SpotsTotalQualityManagementwww.balluff.com

Page 5:6Identify Trouble SpotsStep no. 1In the world of error proofing, any manual operation has the potentialto induce human error. This includes any manual assembly operationand manual machine setup tasks. These areas require evaluation toensure they meet appropriate quality standards such as Total QualityManagement (TQM). As a general rule, every manual assemblyoperation should have at least two Poka-Yoke checks.Evaluating colorOver a short period of time, anoperator becomes desensiti-zed to colors. Color sensorsare proven reliable in situationswhere repetitive color selection isrequired.Installation sequenceRepetitive installation se-quences are often identifiedas troublesome areas. Mostcommon are operations wheremultiple parts such as washers,spacers, O-rings and fasteningdevices are stacked on a com-mon component.Missing or incomplete threadsThreads are a common problemin manufacturing. Missing orincomplete treads can occur withbolts, studs and tapped holes.Missing or wrong partsMany parts are small with littlevisible distinction. Operatorshave trouble if systems are notin place to detect the wrong ormissing parts.Some common trouble spots include:✓ ✗ ✓ ✗✓ ✗ ✓ ✗ContainDiscrepancies ImplementDeterctionIdentify Trouble SpotsTotalQualityManagement

Page 6:7Products being mislabeledor mismarkedAnytime there is manual markingor labeling, there is potential forerrors. These possible errorscan be avoided with a simpletraceability system.Critical fastener torqueand sequenceElectric torque controllers area step in the right direction,but issues can still arise. Themain problem is in knowingthe exact bolt that a specifictorque is being applied to andthe tightening sequence. Spatialpositioning for torque tools hasproven effective to handle thissituation.Manual machinesetup problemsMistakes with manually switchedover machine change partsor manual entered configura-tion data can cause immediatemachine malfunction, but worseyet are the ones not caughtimmediately. These errorscan make their way down theprocess, compounding the errorcosts. Automatic identificationwith RFID or bar code is a simpleand effective means to ensurethe machine is set up correctlybefore the machine goes backinto production.Manual data entryThe average error rate for keys-troke entry is 1 %, or one errorin every 100 keystrokes. Thebest method to prevent errors isto eliminate data entry by usingRFID or bar code systems.GLUE DISFEED: 12PRESSURE: 18TYPE:T6_ _ _ _ _ _ _ _RECIPE: B✓ ✗1 34 21 34 2✓ ✗✓ ✗RECIPE: TOOL BA B C✓ ✗✗www.balluff.com

Page 7:8Discrete sensors provide an on-offsignal to indicate non-conformance.Analog sensors take an actual measurement ideal forhighly flexible applications or statistical process control.Implement aDetection MethodStep no. 2There are two types of error proofing used in manufacturing –passive and active.Passive error proofing uses mechanical keying that ensures aprocess cannot be performed incorrectly. This is an effective andeconomical option; however, it’s not very flexible.HiLoHiLoImplementing sensorsSensors provide standardized outputs that are either discrete(yes/no) or analog (measurement). Which one to use depends onthe level of error proofing needed. Discrete sensors are simple andextremely easy to integrate. Analog sensors are able to conveyactual measurements or product position information. Either of theseoutputs can interface directly to a modular expandable I/O systemthat interfaces to the lockout or rework diverter functions. Simpleindicator lights, panel meters, or a man machine interface can alsobe used.ONOFF✓ ✗✓ ✗ContainDiscrepancies ImplementDeterctionIdentify Trouble SpotsTotalQualityManagementActive error proofing uses sensors and vision systems to activelyverify that a process is completed correctly. Active error proofingis much more flexible and can provide more data than a passivedevice. Furthermore, traceability is easily integrated with active errorproofing. Implementing active error proofing is accomplished byusing either a discrete or analog sensor, color sensor, or a visionbased sensor. Sensors are simple and cost effective, while visionbased sensors are capable of more detailed inspections.

Page 8:9Implementing VisionVision based sensors take over where traditional sensors leave off.Vision can be generally divided into two basic categories: VisionSensors and Vision Systems. Vision sensors such as Balluff’s BVSline are ideal for error proofing applications. They are simple, costeffective, and flexible. Vision systems are designed for complexoperations such as robot guidance, image analysis, and imagecapture and storage.Passive vs. active error proofingParts on free conveyor Detecting randomly located nutsDetecting multiple pins in a connectorParts are not well fixturedIf the part is not contained in a fixture, orthere is no opportunity to bring the partinto an inspection station that has bettertolerance, then a vision system is the bestchoice. Example: parts directly on movingconveyor belt.Multiple inspectionpoints per partIf there are multiple detailson the part to error proof,vision systems are recom-mended. Example:inspecting multiplepins in a connector.Location of detail isnot known or is randomIf the location of the detailon the part in question isnot constant or its locationis random, then vision sys-tems are an ideal choice.Example: random locationof bolt on stud.www.balluff.com

Page 9:10Sensor BasedDetectionYes/No – discrete sensorsSimple and effective, discrete sensors provide yes/no results for manyapplications. These sensors are ideal when parts are fixtured well.Thread detectionStandard diffuse laser is used tolook for the threaded surface.Trim installStandard diffuse laser is usedto see the shade and/or texturedifference of an install trim piece.Part positionLaser retro-reflective sensoris used to ensure fastener iscompletely seated in place.Missing partLaser though-beam is usedto ensure critical componentis in position.Missing partBackground suppression laser isused to detect small componenton close background.Installation sequenceStandard diffuse laser is used to see the shade differencebetween the metal washer and the black sealing washer.ContainDiscrepancies ImplementDeterctionIdentify Trouble SpotsTotalQualityManagement✓✗✓✗✓✗✓✗✓✗✓✗

Page 10:11Measurement sensorsBy using a measurement sensor,the correct part or componentcan be verified. This also aids inflexible manufacturing.Size verification –Part seatedWith multiple fasteners andpossible depths, a measurementsensor can provide all the detailsneeded for verification.Tool position – LinearSpacial position applications with linearmovement are reliably and economicallysolved by linear position traducers.Tool position – RotarySpacial position applications withrotary movement are reliably andeconomically solved by magnetictape encoders.Sealant/grease detectionWhen UV tracers are present inthe sealant or grease, detectionof proper material and amount iseasily accomplished.Color matchingThe color sensor is taught aknown good color and is thenready to detect that color basedon a tolerance factor.✓ ✗ ✓ ✗Sensors for comparing position, color and materialWhen objects need to be scanned and have their positionchecked, you need sensors for position and distance measurement.Photoelectric sensors detect the right color and material.www.balluff.com

Page 11:12The Perfect Error Proofing ToolVision simplifies complex sensing applicationsBalluff Vision Sensor series application examplesThe Balluff Vision Sensor can replace many different discrete andanalog sensor functions along with entire sensor arrays in a singleerror proofing device.Automotive (BVS Standard Version) –Dimensional gauging/component inspection:Inspect fuel injector for verification/presence of electrical cap,injector nozzle, o-ring, and gauging for the proper injector width.Automotive (BVS Advanced Version) –Geometric pattern matching searches X, Y, and360° rotation for geometric features:Verify presence of bearings and size of gaps.Edgeposition WidthGeometricpattern matchWidthIn most production situations, vision systems can be overkill – tooexpensive, too much functionality, and just too complex. Instead,Balluff vision sensors are easy to set up, simple to use, and quickerto return your initial investment.The Balluff vision sensor is a powerful error proofing tool that can beused in almost any area of your manufacturing process. It providesreliable part or feature presence/absence and position detection,plus dimension verification and accurate barcode reading withcrisp and reliable resolution. The Balluff vision sensor has far morefunctionality than any discrete sensor, sensor array, or vision productin its class.Increases product quality■■ Eliminates unreliable manual inspection■■ Allows 100 % quality checking instead of audit checking■■ Provides the resolution needed for reliable quality inspection■■ Enables automated barcode readingReduces costs■■ Single-unit operation replaces expensive, cumbersomemulti-sensor solutions■■ Four models with multiple performance levels to choosefrom provide multiple price points based on functionality■■ Single easy to use software package minimizes setup timeand cuts startup costs■■ Provides vision performance at smart sensor pricingIncreases productivity■■ Improves line speed and error proofing by eliminating the needfor manual inspection■■ Minimizes false code reads with very high code resolution forgreater reliability■■ Catches errors sooner to reduce unplanned downtime and scrap■■ Reduces planned downtime with greater functionality and flexibilityVerify presence of componentsand proper injector widthMissing o-ring,missing cap Bearings present, gaps correct Incorrect gap✓✓✓✓ ✗✗✗✗ ✗✗✗ContainDiscrepancies ImplementDeterctionIdentify Trouble SpotsTotalQualityManagement

Page 12:13Electronics (BVS Standard Version) – Assembly verification:Inspect a circuit board to ensure that all connectors are present andverify the correct orientation of all the capacitors.Pattern matchPackaging (BVS ID Version) –Verify barcodes■■ Codes are read irrespective ofslight changes in position■■ High detection rate■■ Part identification■■ Process monitoring■■ Print monitoringPackaging (BVS ID Version) –Verify data matrix codes■■ Codes are read irrespectiveof the position■■ High detection rate■■ ECC 200 compliant■■ Part identification■■ Process monitoring■■ Print monitoringBarcode Data matrix code✓ ✓✗All connectors are present andcapacitors are in correct orientationBad orientation,missing connector✓✓✗✗ ✗✗✗www.balluff.com

Page 13:14ColorSequencingWith flexible manufacturing, product color is constantly changingon the same production line. Over time, humans have difficultyidentifying color. This fact has driven true color sensors to becomevital Poka-Yoke devices. These devices are simple and effective.Just teach all possible colors and then the sensor will signal if yourintended specific color is present.To address sequencing challenges, true color sensors verify thecomponents when preparing the shipment as well as verifying theshipment once it has arrived. When multiple color combinations arerequired, true color sensors match the sub-component color to themain component color signaling a match.Color matching Color shade verification✓ ✓✓ ✓✗ ✗ ✗✗ContainDiscrepancies ImplementDeterctionIdentify Trouble SpotsTotalQualityManagement

Page 14:15Traceability forJust-in-SequenceJust-in-Sequence (JIS) is a subset of Just-in-Time (JIT) materialflow. Common in flexible manufacturing assembly lines, JIS not onlybrings sub-components to the line at the right time, but also in thecorrect sequence. By sequencing the sub-components to match theassembly sequence, critical automation equipment always pull thecorrect parts.Traceability is the method by which multiple vendors can ensuredelivery of sequenced products. By using Ultra High Frequency(UHF) Radio Frequency Identification (RFID) in combination with amulti-barcode scanner, part shipping pallets and containers canbe tracked with exact pocket location. The specific build data onthe part from the 2D barcode can be combined with other partsinto a database logged with the RFID tag. By using RFID, vendors’systems do not have to be highly integrated to achieve error proofsequencing.www.balluff.com

Page 15:16Traceability is an integral part of containment. In flexiblemanufacturing environments, traceability data identifies and tracksthe specific version to be manufactured. This is accomplished withbuild data, which is a complete description of the intended part. Forexample, if the build data calls out a dark blue part, the color sensormust confirm this. Thus, the Poka-Yoke devices verify what the builddata is asking for.Discrepancies occur when build data and Poka-Yoke device datadoes not match. There are 3 main ways to deal with discrepancies:■■ Scrap the part: Containment is a major issue. Care must betaken to immediately destroy the part or mark it in some way thatit cannot be used downstream. Use of a traceability system tocontain the part is essential.■■ Reclassify the part: In very select applications, products notmeeting eligible parameters can be reclassified into different pro-ducts. This type of application requires tight traceability controls.■■ Rework the part: One method of rework is to stop productionand fix the part on the spot, in station. However, this can haverepercussions throughout the plant. A better method is to incor-porate a rework area. Rework areas are common when sequentialconveyors are used to move product through various assemblyoperations. When an error is detected, downstream processescan be bypassed and the repair technician is shown the exact na-ture of the error. After repair, traceability data allows the reworkedpart to start over again, skipping steps that have already beenperformed.Contain Product DiscrepanciesStep no. 3ContainDiscrepancies ImplementDeterctionIdentify Trouble SpotsTotalQualityManagement

Page 16:Take control of your rework areaReworking a nonconforming part is challenging. Rework areas oftenhave limited controls making it possible for a part to enter with oneproblem and leave with a new one. Fortunately, traceability enablestighter controls.Based on the actual problem with the part, constraints are placedover the rework area to limit the access of rework technicians. Thiscan be done by locking out certain tools using spacial positioning toconstrain tools to the affected areas. Spacial positioning monitorsthe tool’s actual position using linear or rotary position sensors toensure the correct area is being worked. More common are electrictools that are simply disabled if in the wrong position.When rework is complete, traceability information will guide the partback through existing in-process inspections as deemed appropriateby the process. These records are also maintained for future trainingand process improvements.Spacial positioning of a torque wrench17www.balluff.comLinear and rotary position

Page 17:Add Complete Visibilityto your Assembly SystemsRFID enables complete traceability inflexible manufacturing systemsManufacturing plants utilizing automated assembly have uniquerequirements when it comes to tracking their Work-in-Process (WIP).Accurate, real-time tracking brings complete visibility to processes.Tracking can also include all the lineage information from all thecomponents used in the final assembly. Most automated manufac-turing lines also utilize flexible manufacturing where multiple productversions are made on one line. When looking at automated assemblyin total, there are three primary areas of tracking: Build, Process, andLineage Information.18BIS V Industrial IdentificationContainDiscrepancies ImplementDeterctionIdentify Trouble SpotsTotalQualityManagement

Page 18:19www.balluff.comTraceability – A key to a strong containment strategyWhen looking at traceability in assembly plants, there are three primaryareas of tracking: Build, Process, and Lineage Information.Build information: Used heavily in flexible manufacturing environ-ments, build information fully describes all variables in producing aspecific part. It is essentially a build sheet that instructs the assemblyprocess what is required to produce that unique version. Build infor-mation can be held locally on an RFID tag or in a centralized databaseidentified by the tag.Process information: With dual uses, process information is critical tothe manufacturing process. It holds the results from all the in-processtests and Poka-Yoke devices. There are two main uses; flow controland archiving. With flow control, in-process test results are conveyeddownstream so the process flow can be adjusted accordingly. If aparticular part failed a test, downstream processes can be bypassedand the part can be sent to a rework area, where the exact problem isthen presented to rework technicians. Archiving is used for postpro-duction tracking. Process data is stored for later use in recall, liability,and regulatory situations.Lineage information: Similar to archiving process data, lineage dataadds the additional step of tracking each component used in the finalassembly. By consolidating all component data with the process infor-mation, a part’s complete construction is documented. This is vital inthe event of a product recall, and to maintain regulatory compliance.

Page 19:20Your preferred networkSupporting SystemsArchitectureControl architectures have significant impact on how a total errorproofing and traceability program is implemented. Having a systemthat works well initially is somewhat easy. The real test is how wellit works as time goes on. In an error proofing environment, easyexpandability is a unique requirement.It is important to specify the appropriate control architecture duringthe initial design and build phase. This requires an architecture thateasily integrates Poka-Yoke devices and traceability devices such asRadio Frequency Identification (RFID) into one seamless system thatallows easy and low cost expansion for the future.Balluff has developed modular expandable architecture built aroundthe open standard of IO-Link. This architecture seamlessly integratesPoka-Yoke devices and industrial identification devices. By keepinga few IO-Link ports open, future expansion is easy and cost effecti-ve. And the best part is the ability to implement the modular expan-dable architecture on popular control platforms from Allen-Bradley,Siemens, Mitsubishi and others.ContainDiscrepancies ImplementDeterctionIdentify Trouble SpotsTotalQualityManagementPoka-Yoke devicesVision and bar codePneumatic gate controlPressureRead only RFIDRead/write RFIDPosition sensingLaser sensorSensorsProcess visualization