Laboratories face increasing pressure to automate their operations as they are challenged by a continuing increase in workload, a need to reduce expenditure, and a demand for accuracy and turnaround time—essentially doing more with less.

As a global reference laboratory operating within Mayo Clinic’s Department of Laboratory Medicine and Pathology, Mayo Medical Laboratories (MML) performs approximately 23 million tests annually for more than 4.5 million patients worldwide.

MML is comprised of eight laboratories: Anatomic Pathology, Clinical Biochemistry and Immunology, Clinical Core Laboratory Services, Clinical Microbiology, Experimental Pathology and Laboratory Medicine, Hematopathology, Laboratory Genetics and Genomics, and Transfusion Medicine. Each laboratory tests for a wide range of diseases, treatments, and services.

With so many tests, it’s imperative to have a system in place to ensure each and every test is performed accurately and in an efficient manner.

MML implemented a sort automation system in 2014 to track each test, identifying the intended laboratory and accompanying test details. The sort automation system allowed MML to ensure continuous testing operations while improving turnaround time, decreasing laboratory errors, and increasing accuracy and throughput.

This article provides a thorough overview of MML’s sort automation system, along with top lessons learned throughout the entire process of selecting, building, and implementing the system.

Executive Summary Overview—Selecting the System

Out with the Old, in with the New

In 2011, it was determined that the old sort automation system was getting to the point where it was having some mechanical problems, and based on the volume of growth at that time, it wasn’t producing the speeds MML needed. The company was experiencing repetitive downtime issues and increasing maintenance costs. Instead of trying to “fix” the old system, it made more sense to evaluate new vendors and implement a system that would position MML to continue growing.

Finding the Right Dance Partner

While the system is obviously an important component, selecting the right vendor is just as equally, if not more, important.

It’s critical to select a vendor that you know can provide the system you need but also that you can trust and rely on well into the future.

MML spent nearly two years in the process of selecting its vendor. A team visited several industry tradeshows to get a better understanding of what vendors were doing with high-speed sorting, analyzed the capabilities, and compared them to what the company needed for its own laboratory test sorting. From these shows, the MML team identified a short list of companies, ultimately selecting Yaskawa Motoman, a leading robotics company in the Americas as its partner for the new automation system.

MML narrowed its vendor list by analyzing the following capabilities:

  • One vendor for the entire solution
  • Hourly throughput
  • Track-connected automation system
  • 2-D barcode reading capability
  • Support and maintenance capabilities

Starting the Conversation

After selecting the vendor, MML had a representative from Motoman visit its laboratories in Rochester, Minnesota, to assess its current system. MML also visited Motoman on site at its Irvine, California, facility to better understand its capabitilies, not only from an equipment perspective but also its ability to manufacture a system of this size.

What to Expect during Your First Meeting

During your first conversations with a potential vendor, it’s important to ensure the vendor knows your expectations. Key themes to touch on include:

  • Understanding your own needs/expectations and process.
  • The vendor should be open to accept your challenge and willing to work with you to implement a solution.
  • Having an absolute minimum requirements checklist in hand.
  • Options, options, options. A one-size-fits-all approach will not work as every laboratory has different sorting requirements.

It’s Never “One and Done”

After selecting the system, it’s never as easy as just saying “yes” and watching the installation happen the next day. The MML team tested the sort automation system thoroughly before it made its way to Rochester. The team took several trips to California to see the system being manufactured and to make sure the system passed all required industry tests.

The system had to pass standard test processes, which included installation qualification, operational qualification, and production qualification.

After passing these tests, Motoman also conducted factory-acceptance testing. Upon completion of all testing, MML had confidence in the system’s ability to pass all criteria once installed on site.

System Overview

The Nuts and Bolts

The new sort automation system is comprised of four main parts:

  1. Loading stations
  2. 2-D barcode reading stations/RFID binding
  3. Container routing/transportation track system
  4. Sorters

Essentially, the operator loads the tubes, which are then pulled into the barcode reading/RFID binding station. Each tube is picked up, spun around to read the 2-D barcode, and then “married” to an RFID tag. From there, it is sent out to the main loop. The operator loop moves it to the sorter loops, which then bring it into each individual sorter. The sorter reads the RFID tag, picks it up, and knows where to place it based on its identification. The RFID component is critical because it allows the system to know exactly where the specimen needs to go, much faster and more accurately than manually picking up the specimen and reading the barcode.

System Features

  • High-speed sorting instrument enables pre- and post-analytic specimen processing.
  • System operates as a “standalone” sorter or as one in a series of processing stations in a conveyor line.
  • Configurable deck features drawers and/or flow-through conveyors to feed specimens into the sorter and to present racks for sort targets.
  • Puck-based, modular specimen transportation solution can be configured to meet laboratory requirements.
  • Large sort deck yields extended walkway times.

System Benefits

In comparison to the previous system, MML experienced significant benefits from the new sort automation system, including:

  • Reliability: The old system failed quite a bit or required downtime maintenance, which sometimes caused a specimen to get lost or even damaged. The new system ensures reliability with the ability to track each specimen to its intended location.
  • Throughput: The new system handles approximately 6,000 test tubes per hour with the capacity to expand. The old system could only handle 3,000–3,600 an hour.
  • The “Ooh-Ahh” Effect: When potential clients see the new system, they are literally wowed by the magnitude of the system and its sorting capabilities.
  • Not Completely Custom: The previous system was built to extreme customization to fit MML’s needs. At the time, that was optimal, but, ultimately, it proved difficult to maintain. The new system uses a standard hardware platform that can easily be maintained to ensure continuous operation.
  • Modifications: The system does feature some modifications for MML, such as the sorters, to meet MML’s specific sorting needs.

Lessons Learned

While the new sort automation system is efficient and productive now that it’s up and running, in hindsight, there are definitely steps that could have been done differently to streamline the process. MML put together a “Top 10 List of Valuable Lessons Learned” to help you plan your sort automation system purchase and installation:

  1. There’s no need to rush: It’s important to have everything in place before you just jump in and purchase a machine. MML was trying to get the necessary approvals to make the purchase before even knowing what our requirements were or what we really wanted out of the system. This made it extremely difficult at times throughout the process.
  2. Budget assumptions: We had to make significant assumptions on cost to receive capital funding for the project before we even had a chance to see the final system plan. From a project perspective, we didn’t have adequate time to plan and budget for the exact machine we wanted. In the end, it all worked out, but it could have been less stressful if we had been able to identify the solution and budget accordingly.
  3. Planning, planning, planning: It was chaotic for a period of time trying to manage every detail. For example, items surfaced, such as the spacing in the laboratory, how big the system would actually be, and how to train employees on the system. It would have been beneficial to have all criteria identified first and foremost, and all the players in place.
  4. Communication is key: It’s important to communicate not only within the team and department but also across other departments. Regular project meetings are valuable so everyone involved has insight in the project. Also, weekly updates to administrators, managers, and supervisors are helpful.
  5. Gaining support from the get-go: Gathering necessary support from key leadership is beneficial throughout the entire project as it helps to streamline decisions and ensures everything keeps moving forward.
  6. Have a risk-management plan in place: Even with thorough planning, there were aspects that just couldn’t be tested or assured before installation. It’s crucial to have a risk-management plan just in case something does not go according to plan.
  7. Make friends with IT: With a system of this size, and software in play, there’s always the chance that something might not work as expected. Having support from IT ensures help is only a phone call away.
  8. Address problems face to face: Instead of addressing problems by email or phone, it’s valuable to address problems face to face in a “war room” working the issues, real-time.
  9. Break it until you can’t anymore: In this case, Motoman encouraged MML to spend a lot of time on site, essentially breaking down the system and the processes to find any and all glitches. This ensured that when the system arrived in Rochester, it was nearly spot-free.
  10. Have realistic expectations: When the system is in place and ready to go, don’t expect it to run at full capacity on the first day. Implement a slow rollout plan to ensure everyone familiarizes themselves with the system and is trained properly.

Assembling the Team

The project team is at the heart of the solution and requires several individuals, including:

  • Laboratory management
  • Project manager(s)
  • Process engineering
  • Subject matter experts (supervisors)
  • Information technology staff

 

Results

With the new sort automation system in place, Mayo Medical Laboratories has experienced several advantages and benefits. Let’s break it down:

  • Increased throughput: The new system can process 6,000–9,000 test tubes per hour, which is an increase from approximately 3,600 test tubes per hour maximum.
  • No added FTEs: No new full-time employees were needed to sustain testing growth.
  • Reduced failure-testing rate: About 95% of test cancellations were related to the old sorting system. This number has dropped significantly.
  • Decreased mis-sorts: The old system mis-sorted approximately 20–30 specimens per day. To date, there are zero mis-sorts with the new system.
  • Aesthetics: When new and potential clients visit Mayo Medical Laboratories, the new system is intriguing, admired, and appreciated.
kschrib

Kelley Schreiber

Kelley Schreiber is a Marketing Channel Manager at Mayo Medical Laboratories. She is the principle editor and writer of Insights and leads social media and direct marketing strategy. Kelley has worked at Mayo Clinic since 2013. Outside of work, you can find Kelley running, traveling, playing with her new kitten, and exploring new foods.

Responses

Very informative and impressed with the capabilities at Mayo medical Laboratories. Thank you for sharing the news.

Interesting article Kelley, thanks for sharing. We have had an automated line at our hospital in Ithaca, NY, for about 4 years now, and it’s a great help especially with add on testing. One of the issues we have from time to time is the inability of the bar code readers to work when the specimen label is even slightly off. It seems that there is a relatively tight window where they are able to read the barcode, which then holds up the system until it is fixed. I was curious, with a automated system of your size, if you have this issue. Obviously, it would be much more problematic with your volume.

Hi Greg, thank you for your comment. According to Josh Lee, Supervisor in Operations, we do not have any issues with our system’s ability to read barcodes. Based on Josh’s experience, the camera’s field-of-view can be adjusted; however, if your system requires that the specimen remain in the transport mechanism (i.e. puck) during the scanning process then label placement will play a major role in the system’s success. We had a similar issue with our previous system, however, in our new system we are picking from the transport mechanism to scan the specimen, which reduces the constraints on label placement.

Comments are closed.