Conveyor Design

At work, I was asked to look at a conveyor design to see if I could improve the reliability of the system. After looking at it a while with other engineers commenting in, we never could come to a complete agreement on what was the best idea.
For fun, I decided to ask the populous what designs they have seen all over the world. I would really like to hear what you all would say about the system or even systems see in the past that work well or don't work at all.

The system I was asked about.

Problem:
Multiple conveyors are used for transferring products to different areas of the plant. Any downtime results in major losses in the product. The number one goal of this system is to keep running at all cost and to track each product using barcodes on the products.

Solution 1:
Create a system with redundant an Allen-Bradley PLC and remote I/O controlling each conveyor section. The network and communication are done using the loop topology keeping it exclusive to the conveyor system. An automated barcode reader is used at choke points for barcode verification.
*Single PLC easier to track barcodes
*Only one PLC could reduce availability

Solution 2:
Create a system with multiple PLC racks controlling larger sections. The communication between PLC is done using OPC to keep track of barcodes information. An automated barcode reader is used at choke points for barcode verification.
*Multiple PLC means sections could go down while other lines are running. Less loss.
*OPC provider is necessary for barcode tracking which creates more overhead.

Lets see what you all have.

I would vote for a single (Redundant) PLC rack, it would mean less PLC’s that can fail.

Breaking the conveyor system up into subdivisions and having a PLC controlling each section increases the likelihood of having a PLC failure on any given day, and thus bringing down part of the system. Can the entire system run with any other section down?

I have seen far less PLC failures then motor, gearbox, barcode reader failures. And then there is the added complexity of communications if data needs to be shared between the lines.

I think you need to look closely at exactly what causes these failures. I seriously doubt that they can be traced back to the PLC unless the original installation was severely bungled! It is much more likely to be external devices that fail and no amount of PLC magic will fix those problems.

Welcome to the forum.

What type of conveyor system?
CDLR, accumulating belt, stations per section.
A few more details about the system would be helpful.


Our biggest problem is bearing failure, motor failure on the conveyors.
there is not much you can do about that.

regards,
james

RhynoEng said:

Problem:
Multiple conveyors are used for transferring products to different areas of the plant. Any downtime results in major losses in the product. The number one goal of this system is to keep running at all cost and to track each product using barcodes on the products.

Click to expand...

If all of the conveyors are part of one line (e.g. conveyor 1 feeds into conveyor 2, and so forth, as is done in an automotive plant), then breaking things up into multiple PLCs isn't going to help much when one section is down.

Other details that you should provide if you want a reasonable answer:

How many conveyors are we talking about?
What areas will they be going through? Ovens? Tanks? Paint? Anything that would disrupt the barcode?
Do the carriers/packages need to be kept in FIFO order?
Are you going to have to replace your barcodes at any time during the process?
Etc...

As for the barcode reader, I prefer to have them tie directly into the PLC. If you have ever had to deal with the IT folks, and with what they feel passes for PLC ladder logic, you would understand.

No need for a redundant PLC.

RhynoEng said:

Lets see what you all have.

Click to expand...

Experience.

I agree with mendosey. I think you need more information on the types of errors that you are having. Break each error down and understand the exact cause of the error and corrective actions taken. A fishbone diagram can be useful to organize your thoughts.
https://upload.wikimedia.org/wikipedia/commons/thumb/5/52/Ishikawa_Fishbone_Diagram.svg/280px-Ishikawa_Fishbone_Diagram.svg.png
Regards,

PLC isn't going to be the problem for MTF's. IMHO a redundant PLC is a waste. Use the money for better motor/gear boxes, or quick change spares or alternate routes in the case a conveyor goes down.

Agree

robertmee said:

PLC isn't going to be the problem for MTF's. IMHO a redundant PLC is a waste. Use the money for better motor/gear boxes, or quick change spares or alternate routes in the case a conveyor goes down.

Click to expand...

I have seen this used many times where conveyor down time is an issue. They use dual conveyors side by side with crossovers so when a section goes down, everything crosses over to the other section and goes around the down section.

So I think most agree that its going to be mechanical failures (motors, bearings, belts, sensors, etc) not the PLC... I have done some conveyor systems that were modular in design, so if you had a issue you would pop out a section and put in the spare then your up and running again, each module was about 20ft so storage was a little bit of a pain but thats why the warehouse people did not like us

We design material handling systems in massive distribution facilities for many major retailers.
I can tell you that PLC failure will be the very last of your problems.

Redundant PLCs are pointless. All you need is spare parts and program backups.
AB PLCs have been running for 20+ years without a hiccup.

What you need is parts for every conveying unit in question... organized well so they can be found quickly.

All conveyor units should be identified with labels, parts on unit should have easily identifiable part numbers.
Most parts in the system can be replaced in under an hour if everything is organized and prepared well.
This is how all the major players do it.

Aside from that redundant product paths would be my next recommendation.

Avoid OPC

Agreed - the PLC won't be where you get your breakdowns.

Agreed - spares warehoused for everything accessed easily is likely the best way to reduce downtime.

Preventive Maintenance on the bearings - a hand-held infra-red gun will show bearings on the way out. It takes a walking tour of a tech once a week and an excel sheet to trend. Online data is better but much more expensive.

Preventive Maintenance - vibration on the bearings (motor and gearbox) would be the next step. You'd still need the thermal gun for a roll conveyor or the rollers on a belt conveyor.

Auto-greasers on the important bearings - mostly gearboxes - would be my next step. Fill them once a day/week/month and plug them in, let them keep everything greased.

Level switches on the gearboxes, if they use oil lubrication instead of grease, is next.

Monitoring the amps on the motors and the temp of the gearboxes is likely not worth the cash. We only do that on 250 HP and above. The motor amps go up over time as the motors lose some insulation rating. The motors heat a bit. It's not useful until you find out how much your average motor current increases before it dies. More long term.

Going with a high resistance grounding system is expensive but will help isolate ground faults on the motors. This is more of a safety thing but it does help a bit on the downtime. Ground fault monitors will alarm before they trip, so you get warning alarms that will let you decide if you want to run through to the next shutdown or do it now.

Bar code readers can be read by the PLC directly. Adding a PC, especially a Windows PC, that does the OPC storage and translation makes things less reliable, IMHO. I have vintage OPC stuff (mostly Matrikon) that shows the wrong status on their communication channels (shows good but is really failed). We have spent a long time getting rid of OPC, which was supposed to be the greatest thing ... circa 2000.

To make the system more reliable I would suggest backup detection of the products on the conveyor. Perhaps capacitive proximity detectors if you are working with roll conveyors, polarized photo detectors if you are working with a belt-type conveyor. Something to tell you when you have a product but the bar code scanner has not reported a good read. This would check for product that twisted on the conveyor so the scanner cannot read it, or dirty bar codes, plastic wrap that reflects the scanner laser unevenly, etc etc. There are many things that cause bad reads on bar codes. And it only takes one bad read in 10,000 to be a pain for troubleshooting.

The PLC code gets more complex. You end up with a bit of a state machine to determine if product has moved from one conveyor to the next properly. The error tracking gets a bit involved. Depending on who is doing the troubleshooting (ie, is it you?) it may be worth the effort. Make your life as easy as you can. And give yourself all the tools that you can afford (scan time, PLC memory, logging)

Logging all transitions plus date and time stamps on all of the switches, motors, amps ... everything you have ... is pretty high on my list. I log every sensor on our systems so that when something goes weird, I have a shot at figuring out if it was actually an electrical problem, perhaps a mechanical problem, or if it was (more likely) operator error and they don't want to own up to it. Once you let operators know that you log how many times they press the reset/start buttons, or reset the overloads on the motors ... they change what they do .. a little bit at least

As usual, this post is too long. But you did ask!

It must be the end of times. I just agreed with every single comment.
I used to install conveyor sort systems for Goodyear back in 1997.