Automation implementation - Where does one start? A checklist.
The decisions behind what to automate and how to go about doing that is often overwhelming. There is the general apprehension about the capital expense of automation. (Note to reader – automation does not have to be expensive. In fact we have seen that it can be very affordable.)
Secondly, there is the concern about stirring up shop-floor apprehensions on the mistaken taboo that automation takes away jobs.
The third factor that can often prove to be daunting is figuring out how to take a manual operation and convert it so that it can be performed by automated mechanisms or robots. Humans, after all, can discern the “shades of grey” and can use their judgment (non-programmable) to decide on whether something appears to be awry and can take the necessary corrective actions. Mechanized systems, on the other hand, can only do the tasks they are programmed to do and alarms out, or worse yet crashes when faced with unforeseen variation or an out-of-spec condition.
Finally, staying abreast with the latest developments in the field of automation (e.g. Cobots, ToF chipsets revolutionizing vision control systems, and so on) is a monumental task. Manufacturing and process engineers already have their hands full with existing system issues leave alone trying to stay up with the latest technology trends. There is the steep learning curve to determine which one or two, from among the plethora of possible technologies are applicable to their problem.
It is because of the above scenario, small and mid-sized manufacturers put off the much-needed transformation of their operations with technologically-advanced, cost- and labor-effective automation solutions. What follows is a checklist that may aid resource strapped manufacturers and operations personnel who have been tasked with implementing automation. Here is a list of various considerations required in an automation implementation initiative.
This checklist is drawn from the author’s practical experience as well as from the review and study of excellent resources that are listed in the References section.
- Is this the right process to automate? Does it meet the three D’s and S criteria? Dull, Dirty, Dangerous, Safety (i.e. need to get human operator out of the equation)? If yes, then proceed. If no, then look for another project that may need automation more than this one.
- Conceptual stage: Sketch out the concept and layout a schematic (don’t go for perfection at this stage - broad brush strokes only) after gathering the process data.
- Consider all existing operations
- Include tooling involved
- Spell out all process elements
- List special requirements, if any
- Perform calculations/gather data on:
- Speeds of movement
- Time every step and determine the required productivity (units/per unit of time)
- The required axes of motion
- Forces and torques that will be involved
- Dimensions of the parts/process
- Sensitivity to dimensional tolerance variations
- Brainstorm ideas
- Copy existing design and replicate minimizing or eliminating the manual operations and control
- Look for cross over ideas from other fields
- Research new ways and techniques. The internet has an amazing wealth of information
- How best can productivity goals be met
- Develop the kinematics of the system at every stage of the process. How to achieve the required movements, forces, etc. gathered in the concept stage?
Consider the following systems. Each has its own advantages, disadvantages and cost implications. The choice will ultimately be determined by which turns out to be the most cost-effective way to achieve the required movements without violating the requirements from the conceptual stages.
- Consider the configuration of manipulators (T.Strzelecki), end effectors and the type of grippers that may be necessary
- What should be considered for the control of the automation system. This decision will be guided by the requirements laid out in the conceptual stage. What is the required accuracy and precision of the kinematic systems in terms of the following? Each significant factor will necessitate certain choices in control system to achieve the desired outcomes.
- Positional accuracy
- Dynamic accuracy
- Vibrations and dampening
- What feedback sensors are required? (Analog/Digital)
- Linear and angular displacement sensors to determine the location of the components of the automation system
- Speed sensors
- Flow rate sensors
- Temperature sensors
- Part presence detecting sensors
- What is the transport system that will be required to move parts from one stage of the process to the next?
- Linear – e.g. conveyors, pick and place arm
- Rotational – e.g. rotary index table
- Vibrational – e.g. vibratory tracks, bowl feeders
- Multi-axis robot
- What is the automated feeding system that will be necessary? Are there any special concerns about correct orientation of the parts?
- Liquids – e.g. continuous or controlled dose pumps or pistons
- Powder or granular materials – e.g. hoppers, belt conveyors, auger
- Individual parts – e.g. magazines, hopper, bowl feed, blow feed
- Is part surface contact damage a concern? Then alternate, non-contact methods may be required, e.g. – electrostatic, magnetic force field etc.
- Consider redesign of parts that are not easily handled through automation. Existing parts may have been designed many years ago when only human operators were a consideration. Some designs are simply not suitable for automated handling.
An essential portion of any automation system is the electronic control system. The electronics part involves coding and logic to make everything work like a symphony orchestra. However, this stage also does follow the age-old adage – “Garbage in, garbage out.” The considerations that have to go into this stage are extensive and we are working on a checklist format of that process as well. The good news is that if you have completed the above checklist steps successfully, then you are within reach of being able to implement the solution.
May the automation journey be successful and rewarding. When done methodically, it usually is.
Dean A. Shafer, Successful Assembly Automation, SME, 1999
Jon Rigelsford, Robotics: Designing the Mechanisms for Automated Machinery, Emerald Group Publishing Limited, 2002
Ross, Fardo, Masterson, Towers, Robotics: Theory and Industrial Applications, The Goodheart-Wilcox Company, Inc., 2011
T.Strzelecki, Control systems of manipulators intended for automation of assembly operations , Mechanism and Machine Theory, Volume 16, Issue 1, 1981, Pages 3-7