Automation Integrator Selection & Project Lifecycle
...or how not to get fired when you lead an automation project
So you want to Automate?
Choosing to automate a process can be a daunting and scary proposition for many manufacturers, even for those that have done it before. These projects can cost hundreds of thousands to millions of dollars and take anywhere from 6 months to 2 years to complete. I explain to people that this is often a bigger decision than purchasing a home (or more accurately having one custom built). You will need to live with the decision and the choices you make for 20+ years. So you want to get it right.
But don’t worry. You don’t have to go it alone. There are hundreds of very experienced and knowledgeable integrators that will help you to be successful. But how do you know which one is right for you?
I’ve given this article as a keynote speech multiple times over the years but never turned it into an article. While this is not a definitive playbook it should give you some ideas, questions to ask, and things to avoid. If you need more help be sure to check out the link at the end of the article to AutomationAMA where we can assist you in navigating this confusing and critical journey.
Phases of an Integration Project
Undertaking an automation project is a two way street. It’s not simply a transaction with a supplier. You, the manufacturer, will need to be intimately involved in the process and it’s success is highly dependant on your input and actions.
The 5 basic parts of a automation implementation are:
Requirement Definitions
Integrator Selection
Project Management
FAT/SAT
Post Mortem / Recap
We shall discuss each of these steps in detail in this article.
Creating Your Requirement Definitions
The first step of any project is to define the problem and the requirements. Once you have determined the process you want to automate you must be able to articulate that process to the integrator. I have a saying “You are the process experts. We are the automation experts.” Without a defined process, parameters, and requirements you are destined to fail. A complete and accurate scope can be the difference between a successful automation project and a waste of thousands if not millions of dollars.
Some of the steps you need to do in this process are:
Define the problem and process - without a clear understanding of how the process is both completed and monitored you have no way of judging the outcome of the machinery.
Create a realistic budget / ROI period - you need to understand how much this is going to cost and what sort of ROI you are seeking. While sub 1 year ROIs are possible in specific cases, anything under 3 years is typically a good ROI. Don’t set unrealistic goals.
Gather required info - this is where you do your homework. You need to gather all the require samples, drawings, work instructions, videos, photos etc.. This can be a lot of work, but integrators need all of this to give you an accurate quote.
Write a clear but concise URS /RFQ - (User Requirements Specification or Request for Quote). While technically two different things (the RFQ references the URS) some people use them interchangeably. I could write an entire article on just this one step (and probably will at some point).
This is the the document that defines what you are asking for from the machine builder. It needs to be specific but not overly constraining. Consider allowing the machine builder to use their prefered suppliers and standards as they have often standardized over a number of years and have better access to and pricing for these suppliers. The URS needs to define things like:Drawings and CAD Models of all SKUs
Physical samples
Videos & Photos (Work Instructions)
Volumes & Takt time of each SKU
Required inspections & quality metrics
OEE Requirements (85% is world class; don’t ask for 99%)
Preferred component list (be willing to bend on this)
Required regulations or rules (FDA, CE, UL , ITAR etc.)
Be prepared to answer questions quickly - if you want a quote quickly, you need to be available and response with answers to questions. Please keep in mind a custom system is a complex project that can take anywhere from 2-6 weeks to quote. Don’t put your machine builder in a position where you need a quote in less than a week. Usually you are going to end up with a higher price than if you gave them the proper amount of time.
Integrator Selection
So you’ve sent our your RFQ to 2-3 suppliers (don’t send it to more than 3, as it’ll be far too difficult to compare them all) and they have sent back proposals. How do you determine which one to go with?
Ask for references and similar projects - ask if the builder has done anything similar before. Ask to see video of the machines or if possible, even to see them in the customer’s facility. Ask for references at their customers’ and ask them how the projects went in terms of both engineering but also project management.
Visit the factory - be sure to go visit the machine builder. You can tell a lot about a company’s process and quality by the state of their shop. I have been into shops where you can eat off the floor and shops where you have to step over junk bicycles to get to your machine (literally). A clean and organized shop is likely and indicator of a company with a process and discipline.
If cost or lead-times are abnormal ask questions - both very high or low prices or lead-times indicate that either the machine builder did not understand the requirements or that perhaps you over/under specified something in the URS. Do not discount the high and low prices but ask them how they arrived at these numbers. It may reveal something insightful about your process or the other quotes.
Read the proposal - actually read it, don’t just flipt to the price page. Make sure they have addressed all the important factors in your scope. Also be sure to see if there are any exceptions or assumptions in their proposal that need to be accepted or addressed. Don’t be afraid to ask for a revision to a proposal if there is something the integrator missed or about which they were not clear. You do not want to accept ambiguity at this point.
Create a scorecard of important factors - it’s often hard to compre quotes to one another, especially if the concepts are vastly different. I suggest making a score card with weighted values (price and lead time will be weighted higher than say proximity). Judge each proposal and supplier on a scale (1-5 or 1-10) and at the end see how they stack up. This is not to say you should choose solely based on this scorecard but it gives you a quantifiable way to compare quotes to quotes.
Be prepared to work on terms or a MSA - a Master Service Agreement is a document that commercially defines the relationship between you and the machine builder. It’s a comprehensive document that cover all projects that you work on together and define tings like payment terms, indemnity, liability, warranty and more. While it’s a lot of work to set up, a MSA, once it’s in place there is no more back and forth about terms, payments, warranties etc.
Also be prepared to work with the machine builder on payment terms. These are often SMBs that are MUCH smaller than their customers. While corporations would love to have net 180+ terms, these machine builders live and die by cash flow and cannot afford to front the costs of all the hardware for 6+ months. You need to work with your purchasing departments to help them understand that these are business that need these progress payments to successfully complete the project. It’s in your best interests to be accommodating (within reason).
Project Management
As previously discussed this is as much of a project for you as it is for the integrator. You will have responsibilities and milestones that YOU need to hit to make the project successful. The more you are involved the better the project will run.
Assign and Request one point of contact - You need to assign a single point of contact (POC) at your company that will be the spokesperson for the project. This does not mean you will not have a team behind you (safety, production, engineering etc.) but there needs to be one person who makes the final call. On the integrator’s side they should provide a single POC as well (typically a project manager). Again you will talk with various people at the builder but you want one central PM.
Set a (variable) meeting cadence - So “variable” and “cadence” might seem incompatible, but you need to set a meeting cadence (say every two weeks to start) to discuss the project progress. However once the machine goes into the procurement phase not much happens while you wait for parts to come in. Be willing to move this to every 4 weeks during this period. The PM at the integrator will thank you! At these meeting you should be given an update on progress, any issues they have encountered, updates on schedule and any milestones that are upcoming that YOU need to meet.
Make sure the integrator has appropriate milestones in their project schedule - While every company is different there should be some basic milestones in the project schedule. These typically include:
Kickoff meeting - introductions to all team members, a review of the scope of the URS as well as the proposal, an initial schedule and setting points of contact on both sides are important aspects
Layout review - a basic CAD review of the rough layout and shape of the machine
Final Design Review - a “speak now or forever hold your peace” review before materials are purchased. This will be a detailed review of the exact design and implementation of the machine. Do not gloss over this.
Assembly Checkin - plan to visit to see the machine once it’s mainly assembled but not yet in debug. This allows you to check on status and see what the machine really looks like. Any changes (ECO or not) are much easier to change at this point (but not as easy as during the design phase).
HMI Review - making sure all the controls and KPIs are present and where you expect them to be. Again it’s much easier to make changes to the screens now than when the machine has been fully programmed and debugged.
FAT - the date when the machine is ready for you to come run it off (more on this later).
Reply to requests quickly. You also have milestones. - As I have stated, you have responsibilities as well. If you owe the builder samples and are a week late this could easily translate into a moth delay on the project. Integrators have multiple projects going on at the same time. If you are delaying them, they will move to the next project and at times, can;t back to yours until a later date. Be sure you are on top f your responsibilities so that the project stays on schedule.
Be prepared to make decisions on scope vs dollars - at some point in the project an issue will arise. Perhaps the production parts provided are not the same as the samples, or the design of the machine has uncovered an unforeseen failure mechanism. The way to solve this might be to add additional inspections or mechanisms etc. The dreaded “change order”. If the URS clearly spells out what is required then push back on these change orders. However understand that there are times where no one could have foreseen the issue and that it’s unfair to expect the machine builder to cover these changes. Determine if the change is a “need” or a “want” and act accordingly.
Involve engineering, operations, safety - Be sure to have everyone with skin in the game involved. Far too often safety is not brought int the project until FAT. At that point it’s too late to make any changes without significant cost and lead times. Had they been involved from day one, allowances and changes could have been made. Be sure to include operators in the process as well. They often know more about the product that engineering does. They will give you insights that can make the difference between a machine that everyone loves and the one they call nasty names…
FAT & SAT
FAT (Factory Acceptance Test) takes place at the machine builders’ facility. The SAT (Site Acceptance Test) takes place at the manufacturer’s facility. The SAT is typically a repeat of the FAT checking any items that were on the punch list however there are always small exceptions.
One of the most critical aspects of a FAT is to ensure that the machine performs to spec before it leaves the factory. You do NOT want a machine that is not ready to ship. While management may push for you to get the machine into the factory, if it is not ready, do not accept it.
FAT criteria should be defined before FAT - I cannot tell you how many times a URS has NO mention of FAT criteria and then we were given a 75 page FAT document two days before the FAT. This MUST be discussed and agreed upon well before the machine is ready for FAT. A good integrator will have already run through your FAT before you arrive to ensure it will pass. In fact if we did not have a FAT document many months before the FAT we had a clause in our contract that we could charge more and delay the machine. It is that critical.
Test all products. Test all failure modes. - Do not fall prey to testing a few SKUs and then assuming the rest will work just fine. Be sure to test all the products that you plan to run. Also be sure to test all the failure modes. If the proper placement of an o-ring is critical for example, introduce failed products (a.k.a. red rabbits) into the system to ensure they are caught.
Bring your operators. They will be invaluable. - I always suggested that the customer bring the operators that were going to run the machine to the FAT. Not only will they gain valuable training they can make small changes to the machine that will affect them for a long time to come. Also when the machine hits the floor they are already familiar with the equipment and have an elevated sense of comfort. Make sure they name the robots too.
Account for enough material for FAT - while materials are expensive and often cannot be sold after runoff, you must have enough to prove out the machine. I see too many manufacturers cheap out on this step and don’t provide enough parts to adequately do a FAT/SAT. Rerunning parts is not always ideal as parts can react differently fresh from the mold vs having been in a feeder bowl for 10-20 cycles.
Understand there will be punch list items - Nothing ever goes PERFECT on a FAT. If there is a critical miss (inspections don’t work, excessive downtime, cycle times not met) then of course you should expect the integrator to fix this and run the FAT again. However minor changes such as changing HMI screens, adding a sensor or a bracket here and there etc. should be allowed to be checked at SAT.
Post Mortem & Review
Once the machine is installed and SAT is performed most people assume the project is over. It’s not. You need to review the process and outcomes with your integrator not only to ensure everyone understands the outcome but to improve the relationship and the next project.
Proposal vs Machine Outcome - discuss anything that was different than what was proposed. In most cases this would have already been discussed at the design review but it’s always good to go over any variances or changes.
Pricing & Leadtime Discussion (ECOs & Delays) - if there were any delays in the project timeline or changes to pricing (ECOs) discuss why these occurred and how to prevent them in the future.
Improvements to next design / project - discuss what could be done differently next time to improve the outcome, whether that is in terms of project flow or mechanical / electrical / controls design. Many times these discussions will lead back to retrofitting the original machines once the improvements have been made on future machines.
Discuss service plans and spare parts - if the integrator has not already provided a service contract or spare parts list, request one. There should be a plan in place in the event that you have unexpected downtime. Also do not cheap out when it comes to spare parts. A single critical item can sometimes have a 20+ week lead time. Try to explain to your management why your machine is down for half a year because you did not buy a $2000 spare part.
Give praise where due - Remember that fundamentally the machine builder is made up of people. If they did a great job be sure to let them know. Many people in this business have a lot of pride in their work and when they are given kudos it goes a long ways.
Conclusion
Your first (or any) automation project can be a scary and daunting proposition. You’ll likely be afraid to make these large decisions that can involve large amounts of capital and time. However if you treat this purchase like any other project, breaking it down into sections and making sure each one is completed, you will be successful. You need to hold the integrator accountable, but at the same time understand that they are people as well. This is a partnership and you are both accountable for the success of the project.
The Automation Navigator is brought to you by Automation AMA, a company founded and run by Sean Dotson a 25+ year industry veteran in assembly, robotics and packaging. Automation AMA offers a variety of services to help you on your automation journey or to help fine tune your engineering, sales, operations and business development departments. Learn more at www.autoamationAMA.com or shoot me a introductory email.