Posts from the ‘Machine Vision’ Category

TPM – Total Productive Maintenance – A DFSS approach?

Here is an interesting link to TPM – Total Productive Maintenance.

What is in essence TPM.

TPM is eliminating 6 types of capacity losses

  • Breakdowns
  • Setups and adjustments
  • Reduced speed
  • Minor stoppage
  • Defect and rework
  • Start up loss

In essence TPM should be a part of the equipment design review during the build process in a DFSS initiative.

There are times in the development cycle of custom machine design where we may be more interested in speed and robustness of process and may not dwell in length on SMED philosophies. So, here again SMED should be a part of DFSS.

100% Machine Vision Inspection will help us detect and correct defects at the source.

So the next time you go and write the URS for a Custom Build Project think TPM along with DFSS.


Lessons for Modern Day Engineers – Be a Master of more than One and Jack of Many

A modern day engineer has to be versatile.

Quoting Robert Anson Heinlein

“A human being should be able to change a diaper, plan an invasion, butcher a hog, conn a ship, design a building, write a sonnet, balance accounts, build a wall, set a bone, comfort the dying, take orders, give orders, cooperate, act alone, solve equations, analyze a new problem, pitch manure, program a computer, cook a tasty meal, fight efficiently, die gallantly. Specialization is for insects.”

For example a Mechanical Engineer in a Manufacturing Company has to be versatile in CAD, PLC Programming, exhibit a high level of computer literacy, know programming languages, understand Servos, Motion Control, Design for Sigma principles, LEAN manufacturing, LabView, Machine Vision, Project Management, good presentation and people skills, be able to make capital requests and understand budgeting.

It is a competitive world and companies prefer engineers who can wear many hats. It is also professionally enriching to cross train on other disciplines as it makes for a very interesting journey and more control over ones projects.

There is so much free information available nowadays in the world wide web and it is very easy to get the required training at little or no cost.

Engineers can also empower themselves by using new techniques and principles in their day to day projects and employers will actually encourage it as long as regular project continue to move.

We are it!

Remote or Site Trouble Shooting Modern Automated Manufacturing Equipment – Introduction

Remote or on site trouble shooting of modern automated manufacturing equipment is fairly straight forward if one follows a structured approach.

You will need to have a functional understanding and access to

  • Human Machine Interface (HMI)
  • User & maintenance manuals
  • Electrical schematics
  • Hydraulic schematics
  • Pneumatic schematics
  • PLC ladder logic
  • Principle of operation and setup of sensors
  • Spare parts inventory
  • Functional modules of the system
  • Process flow
  • Machine vision systems
  • SPC
  • Have mechanical & electrical skills
  • Be computer literate

Trouble shooting methodology starts with problem definition

  • System has stopped with a fault condition
  • System is running but producing excessive rejects – Got to SPC screen and center process
  • System does not start – Check for input power and air, blown fuses.

System has stopped with a fault condition

This is one of the most common faults

  • Look at the alarm screen on the HMI
  • Check cause and then clear alarm
  • Find the alarm which does not clear

Alarm does not clear

The HMI will give the functional input required to be in an on/off state for the system to go to the next sequential state of operation

These could be

  • Guard doors not closed
  • Processing modules not ready
  • System stopped due to excessive rejects – Check reason for excessive rejects
  • Servo faults – Check servo driver, any mechanical obstructions and clear fault
  • Communication error – ¬†Reset PLC to see if connection is made
  • Waiting for supply of input components – Refill required supply components
  • Air pressure low – Go to source
  • Vacuum pressure low – Go to source or check for blockages
  • Input voltage low – Go to source
  • Functional module not turned on – Turn on
  • Scrap bins full – Empty

Guard doors not closed

  • Step by step trace the guard door which is not closed
  • If alarm is still on check for loose fit condition of guard door
  • Check for guard door lock malfunction
  • If necessary use key to simulate close condition of guard door to check function

Processing Module not ready

  • Inspect the functional module
  • Trace the input the functional module is waiting for to continue
  • If it is a sensor check the function of the by simulating an on off condition and check the state in the ladder logic, HMI or indicating LED’s
  • Replace sensor if defective
  • Check if HMI fault condition still exists and go over the check again till this module is ready.
  • Check for bent brackets, loose sensor mountings
  • Check for air & vacuum within limits

This will continue with more detailed approaches on the various functional areas

Till then happy trouble shooting and feel free to post your problems

Machine Vision Inspection – A boon for Kaizen, Six Sigma & DFSS

The design of new custom automated equipment and the process improvement associated with it is the forte of manufacturing, process, automation and industrial engineers.

DFSS (Design for Six Sigma), Six Sigma & Kaizen rely on a measurement phase prior to analyzing the data.

Traditional measurement has been done by CMM’s (Coordinate Measuring Machines)

However this task is time consuming, laborious and not Real Time. This is where Machine Vision Inspection System can provide an immediate return on investment.

Machine Vision Systems are

  • Real Time
  • Work with NIST traceable standards
  • Can do 100% ¬†inspection
  • Are in par with CMM’s on accuracy
  • Are not laborious or time consuming
  • Can stop the system from producing defective parts
  • Process can be centered
  • Processes can be trended
  • Can work in tandem with SPC packages
  • Can be analyzed real time or studied offline
  • Can also have analysis tools built in
  • Involve the user in improving the system
  • Inexpensive
  • Compact
  • Scaleable

Predictive Maintenance Using Machine Vision Inspection & Real Time SPC – Introduction

Machine Vision Inspection & Real Time SPC are now part and parcel of most modern custom manufacturing automation equipment.

This is an inbuilt tool which will also help us conduct predictive maintenance.

  • Real time run charts act as a process signature for a given set of parameters and machine condition
  • Base line these signatures in chart form in your knowledge base
  • Develop history
  • Compare these signatures with your daily run charts
  • Variations in the charts will point out to equipment wear or change in incoming material specifications.

Typical types of equipment deterioration could include

  • Tracking of belts
  • Belt wear
  • Tool wear and time for re-sharpening
  • Maintenance for printers, lasers
  • Bearing failure
  • Broken components in the system
  • Sensors at point of failure
  • Servo motors at point of failure
  • Tools failing in vision system

Machine Vision Inspection – Real Time SPC & Process Improvement

Machine vision has become an integral part of modern manufacturing.

The major components are

  • Cameras
  • Lighting
  • Machine Vision Software
  • Statistical Process Control Software
  • Mechanical Fixturing
  • Computer Hardware
  • Calibration Standards

One of the major advantages in Machine Vision Inspection is that we can undertake 100% percent inspection along with practicing real time process control enabling us to look at see that the process is in control as well as center the process.

Machine Vision in an ideal situation has to be incorporated during the machine design stage.

We have to build in open or closed loop feed back system to center the process as well as see whether the process is in control.

We must have a clear understanding of the parameters to be measured using detailed drawings, documented acceptance criteria and reject mechanism.

The camera and software is fairly easy to choose with the excellent products available from major vendors.

Attention has to be paid to choosing the correct light as well as robust mechanical mountings couple with a very defined calibration strategy using NIST traceable standards.

The next stage is validating the process. Structured validation will reveal flaws in the camera software tools and can be corrected at the beginning.

Real time SPC can help the operators monitor the centering of the process and use the open or closed loop function to control the process. Alarms will indicate if the process is out of control which in turn can trigger preventive and predictive maintenance.

Process improvement and trouble shooting for false rejects or accepts is easy with the facility to log data as well as capture images .

For example the systems can be configured to log data from the measured parameters during a run as well as capture images both for passed or rejected parts. This data can be analyzed with Excel or Minitab to look at abnormal conditions. We can also rerun the captured images frame by frame to see if any of the camera software tools have failed. This will in turn trigger corrective action. Excel & Minitab graphs are excellent presentation aids to present gain made in process improvement initiatives.

In fact with the competitive price structure for machine vision inspection products in the market the is no reason to do any manual inspection for mass produced products.