October 15, 2014
"You're a LEAN...MEAN...FIGHTING...MACHINE!" -Bill Murray, Stripes 1981
Classic movie line, and with a bit of manipulation, an apt description for how many labs go about supporting their instrumentation. Lean, is a concept well known to many manufacturering operations and it's core roots are that of continuous improvement, which is of course finding great interest within research labs.
Lean Maintenance is a proactive approach that favors planned maintenance (PM's) instead of the tradiional annual service contract approach. Annual service contracts are a luxury that most labs (large and smal) can no longer afford. Large labs especially, are moving toward the PM approach for severl key reasons;
It's proactive - good PM's can find problems before they become critical, resulting in greater uptime.
It's pragmatic - mananging multiple service contracts can be very time consuming.
It's cost effective - service contracts costs are a budget killer.
A good PM procedure (one that mirrors the OEM's) requires smaller upfront investments that pale in comparison to annual service contracts costs, which are typically 10-15% of purchase price...year after year. PM's often provide so much cost savings that users are only contacting manufacturers when something breaks.
Finally, in an upcoming post I will talk about some specific strategies for Lean Maintenance. For now, just think twice before you automatically sign up of an expensive service contract...there are alternatives.
September 29, 2014
The good people at LabManager.com have been busy lately, with two great articles on microplate readers.
The first, is a brief overview of considerations for cell-based assays (click here).
There is also a more comprehensive report "2014-15 Product Resource Guide: Microplate Handlers and Readers" (click here). Great info including new product releases and a links to most of the key providers.
Like most new instruments, many of these readers come with 12 month standard warranties (parts & labor). Of course, when your new reader comes off warranty (or your old reader)...you know who to contact for PM and validation services!
July 7, 2014
Plate Handlers = Scary
Okay, so this picture has nothing to do with plate handlers but it kinda freaked me out so, I figured if I'm not going to be able to sleep tonite I might as well take others down with me....
Just saw a recent survey from LabManager.com about micrplate handlers (CLICK HERE).
They suggest 6 questions that potential buyers should ask, but I found the list to be a bit...'thin' and was a little disappointed. A couple of additional questtions might include;
1) Storage - does the robot support both stacks and random access storage? Very useful when working with incubations and other time sensitive events. Are the racks removeable? Very handy for offline load and unload.
2) Gripper - Is the gripper electrically actuated or pneumatically actuated? Electric (servo) grippers offer the ability to fine tune plate gets or puts. Pneumatics will obviously require in-house air or a portable compressor. Also, if you are handling multiple containers or lids, the pneumatic gripper will treat them all alike...not so for the electric.
3) Plate Sensing -Does the gripper include a plate presence sensor? If, so is it force-based or optical? Some optical sensors have problems with clear plastics.
4) Safety - Does the plate handler have an E-Stop circuit? Not all do and even though they can be wimpy compared to industrial robots, there's nothing a quite like an OSHA audit when an unsuspecting operator gets wacked upside the head...
5) Software - Does the plate handler support manual arm movement for teaching or do I have to jog to all locations? Turning off servo motors but leaving the encoders powered up makes it very simple to grab the arm and place it in approximate locations, whereupon you can repower and fine tune. Jogging only takes a lot more time. Also, does the plate handler support Windows 7 or greater? IT departments everywhere are cracking down on pre Win7 software...
Just a few of my thoughts, perhaps you have some to add...
June 2, 2014
LabManager MVS Webinar
WARNING - Gratuitous Self Promotion (shameless, actually)
In case you missed it, LabManager Magazine hosted its annual Multi-Vendor Services Webinar last month, featuring industry expert spearkers from Thermo Fisher Unity Lab Services, Perkin Elmer OneSource, Agilent CrossLab Services and The LabSquad.
If you would like to view the presentations online please click here.
April 29, 2014
Problem Solved - The Mystery of The Plate Sealer That Wouldn't Seal
Okay, so several of you have gotten impatient and sent me emails asking for the solution to the plate sealer mystery in my last post. As is often the case, the answer was simple but it was not obvious...a perfect storm of causes, if you will.
Let me cut (no pun intended) to the chase. The major factor was AC power...or lack thereof. Although the measured power coming into the unit was within spec for the instrument (112 VAC), the user also had a JunAir compressor on the same circuit...as well as a large refridgerator and a centrifuge. The combination of the sealer holding temp, the JunAir running almost continuously (due to a slight leak), the fridge compressor running and the centrifuge...centrifuging, resulted in a transient line voltage drop down below 100VAC. This was only measurable using a chart recorder which monitored the true RMS voltage. While the electronic display of the sealer operated and made it appear that the sealer was functioning, the timing of the pneumatics was not correct. The cutting bar would advance but retract before cutting action could take place. Because the JunAir runs quietly, it was difficult to see that it was pretty much always running due to the leak. We believe the fridge compressor kicking on while the compressor piston was the bottom of it's stroke is where the power drain occurred.
Isolating the sealer and the JunAir (fixed the air leak) from the circuit in question resolved the problem.
Chalk one up for the "doh!" list.
April 1, 2014
The Mystery of The Plate Sealer That Wouldn't Seal
"Ruh-Roh Shaggy!" as Scooby Do might have said when faced with a recent problem we encountered with an ABgene ALPS 3000 sealer.
A customer had two units that operated pretty much 24/7. One unit had tape seals while the other had foil seals, and it was not uncommon to switch between the two.
Iniital failure reports centered around a error message "Not Down." This was pretty simple and invovled the front access door switch. Users like to look inside the instrument as it was operating, so they would tape the switch closed. Occaisonally, the tape would expand and allow the switch to change states. A simple closing of the door and press of the e-stop button would reset the error and allow operation to continue. But, that was not the end...
Soon thereafter, additional calls for help were placed as the instrument would no longer cut seals. The cutting bar which is pretty beefy would only slice through a small portion of the edge of the seal, but never fully perforate. The unit in question was pulled for depot repair and of course upon power up, it ran fine. In fact, it sealed over 50 plates without error. The cutting blade was adjusted and another 50+ plates were run without fail. The unit was returned, but just to be on the safe side, we brought along another identical unit to run side by side with the clients unit.
Onsite, the unit ran great for several hours...then same problem. Our FSE's checked air (80PSI from a big honking JunAir compressor), and AC power...fine. The customer's plate were polypropylene and the temp was set at 167C...a little high, considering polypro has a melting point of 130C. Nevertheless, this is temp the customer had been using prior to failure with no issues.
The solution was simple...but not obvious. Care to venture a guess?
March 24, 2014
Win XP - The Bell Tolls For Thee
A Simple Solution For Older XP Based Instruments
Okay, this time we really mean it! Microsoft is getting really serious about killing Windows XP on April 8th, 2014. Don't worry, no one from the NSA will be showing up with a search warrant on April 9th (...but, you never know). Well, what does this really mean for labs with Win XP PC's? Most IT departments have already made the mandate that Win 7 is the preferred OS for new PC's and many manufacturers have been making sure their software is compatible. But, as we all know, some instruments will be orphaned. That is to say, for a variety of reasons, manufacturers will not be updating software of older XP based instruments. Most labs that I have visited are getting around this in the short term by keeping Win XP PC's off the internal network and sneaker-netting data to a central network location via USB keys. Not ideal.
Below is a simple way to get the data flowing without manual intervention and protecting the integrity of the internal network. Note - this is not the only way, but it is a pretty robust solution.
Here's what you will need;
- Windows XP PC - connected to the instrument (usually USB or RS232) and one Ethernet port. Depending upon the age of the instrument, this PC may be a full tower (big) form factor which takes up a lot of bench space.
- Windows 7 PC - with two Ethernet ports, one of which will be connected to the internal network. You could use a laptop if it has both wireless and hardwired Ethernet (but, since many IT groups really frown upon wireless connections to the internal network, this example will only over desktop PC's). If your instrument PC has a compact factor (size) try to select a PC that is similar in size. It will make for a more tidy installation. You will only need one monitor/keyboard/mouse when you are done, so if the new Win 7 PC has these you will keep these and ditch the ones from your Win XP PC (see below).
- KVM Switch - this will allow sharing of one monitor, keyboard and mouse between the two PC's Be sure that your Win XP PC mouse and keyboard are USB. Some older PC's used the old PS2 style connectors. If this is what you have, ditch them and use the keyboard and mouse from the new Win 7 machine.
- Your IT Support Guru - don't go this alone. If you screw it up you will have to involve he/she to get it resolved so better to keep them on your good side. Besides, this is what they do for a living...you wouldn't like it if they tried to optimize your assays, now would you?
1) Place both PC's next two each other. One PC (Win XP) should be connected to your instrument already, just verify that it is working. The other (Win 7) will be connected to the internal network - make sure that it is working and connected to the network.
2) Plug the Ethernet crossover cable into the Ethernet port of the Win XP PC and the open Ethernet Port of the Win 7 PC.
3) Okay, now here is where the IT Guru comes in handy (remember to complement your guru and have Skittles and Redbull on hand). You are going to want to have Big Bang Theory configure the IP addresses and subnet mask of the two PC's to create a simple peer to peer network. Your IT Guru will know which Ethernet port to muck with on the Win 7 PC such that you don't screw up your network connection. BBT will also need to make sure that TCIP is enabled. EXAMPLE ONLY - LET YOUR IT GURU DO THIS...
Computer 1: 192.168.1.100
Computer 2: 18.104.22.168
The subnet mask entry must also be identical on both machines.
The Domain or Workgroup must also match between the two machines
4) Disconnect the monitor. keyboard and mouse from the Win XP pc and discard them. Connect your Win 7 PC keyboard, monitor and mouse to the KVM box. Most KVM devices have two sets of connections that go the separate PC's and just connect your monitor, mouse and keyboard to it. Pretty straight forward. A hardware button allows you to switch between PC's.
5) When finished, it would be preferable to label each PC and stack them on top of each other to conserve bench space.
6) Have your IT Guru look at where you are storing data Files on the Win XP PC. He/She will need to create a new location on the Win 7 PC and change your instrument programs to start saving data at this location. The data now residing on the Win 7 PC hard drive can now be mapped to your server.
1) Many lab instrument manufacturers use Dell PC's as their controllers, probably because they got a great deal on pricing and consistent quality/configurations. I have seen many an IT department specify Lenovo or HP PC's for their internal networks. Although the specs may look identical...strange things can happen. Most manufacturers will not support (or at least spend a lot of time with you) troubleshooting problems on other brand name PC's. Caveat emptor.
2) When maintaining older instruments, the most oft overlooked component is the PC. This is especially true if you have misplaced any critical instrument installation disks or license into. As long as you are planning to use an older PC, it is imperative that you not just back up the hard drive, but actually image the disk (re: Symantec Ghost). In a pinch, you can still replace a bad hard drive in an older PC and re-image the disk.