Complete Customization, Tracking Everything, Without Programming

The following is a case history of a successful UNIFlow LIMS project that was completely customized to fit the customer's unique requirements without programming. The second half of the project was completed by the customer's own lab personnel, who are not computer programmers. This history is told from the customer's perspective.

Our lab provides a high-throughput compound screening service (HTS). We selected the totally-configurable UNIFlow system by UNIConnect (www.uniconnect.com) because our projects are driven by short-term contracts that always present special needs requiring adjustments to the process, and we needed a LIMS that could keep up with the changes without programming. We also wanted UNIFlow because we were determined to track and control everything that affects quality: our volume is steadily increasing, downtime becomes more and more costly, and we have to be able to quickly identify the cause of a problem.

The core of our process was initially developed by UNIConnect, but mid-project we received the training necessary to complete and extend the process ourselves.

To begin defining our process, UNIConnect asked us to provide a spreadsheet that listed the names of the steps of our core process down the left side and the names of the units we need to track across the top, such as various types of samples, plates, reagents, and instruments. The trackable units are referred to as containers in UNIFlow terminology. In the cells of this table we placed a dot at the intersection of a step and a container, representing the use of a given type of container in a given step. We drew vertical arrows connecting these dots to represent the lifecycle of each type of container, called workflows, from the first step where a container is created to each of the subsequent steps where the container is used. Then we drew horizontal arrows to represent the content flows from one container to the next, representing pipetting, robot fluid transfers, or other compositions.

The spreadsheet with its steps, containers, dots and vertical and horizontal connecting arrows presented a comprehensive and orderly schematic of our laboratory process. This served as the focus of our discussions and helped us make the necessary decisions about the steps of our process and what needed to be tracked.

UNIConnect used a program to read our spreadsheet and generate the initial UNIFlow process definition for our process. The UNIFlow process definition is a text file comprised of tags understood by the UNIFlow engine paired with our names for the steps, containers, and other parameters of our process, using our own terminology. We saw that with training we could read the process definition and edit it ourselves.

The process is defined as a set of steps in a simple indented outline format. Each step specifies the containers and other information to be tracked, the workflow of each container proceeding from one step to the next, and the flow of contents transferred from one container to the next. Using only this definition the UNIFlow engine was able to generate all our web input forms, workflows, content flows, and database interactions to track, control, and record our process.

We did not experience the bugs and debugging process normally associated with new software. The system just ran, exactly as specified in the process definition. We could read the process definition directly and verify that it was performing the work as specified. We began entering test data and could watch it flow through the system, and our technicians started to get a feel for how they would work with UNIFlow.

While we were getting acquainted with the process and the input forms, UNIConnect added additional steps to track peripheral sub-processes such as instrument management, reagent preparation, and screens for other departments to record work they perform at our request. UNIFlow was configured to block the use of instruments that had not been calibrated within a specified time period, and to notify a technician by email before the calibration period expired. We also defined escalating notifications to alert managers if the calibration still had not been completed as the end of the calibration period grew near. This ensures that our instruments are always calibrated and ready for use. We also defined expiration dates for reagents so that UNIFlow will automatically block their use if they grow too old.

About half way through, one of our new lab workers was invited to be trained to take over our UNIFlow process definition and become our UNIFlow process engineer. This detail-oriented person has a deep understanding of our process and key issues, is adept in performing our assays and is confident in working with computers and high-throughput lab instruments. We scheduled a visit for UNIConnect to train our staff on how to use UNIFlow and to train our new process engineer.

The training took about 3 days spread out over a week, and included sessions for lab workers, management, and our process development team. The process engineering topics included:

• The UNIFlow process definition language
• UNIFlow database tables
• SQL (Structured Query Language) used by most database systems
• Principles of laboratory process tracking and control.

The training was adapted to our process and specific requirements, providing a great opportunity to discuss key issues and options with our UNIConnect process engineer.

After the training, our new staff process engineer began adding steps and extensions to the process definition to accommodate new requirements. As changes were made, UNIConnect reviewed them to make suggestions and ensure that our process engineer was understanding and applying UNIFlow features correctly. After a few iterations our process engineer felt increasingly confident and before long was leading the effort to shape our process into exactly what we wanted. Soon the process was completed and validated with end to end runs of the entire process. We have been in production over a year. These days we only contact UNIConnect infrequently when we need ideas on how to handle a situation we haven't seen before. UNIConnect calls us from time to time to stay in touch.

Our IT staff is busy, so we opted to have UNIConnect host and manage our UNIFlow system on servers in their high-security facility via the Internet. Our most sensitive data stays on our computers behind our firewall. Response time and uptime have never been a problem. Our license gives us the option to install the system on our own local servers if that becomes necessary.

The next goals for our lab include managing our reagent and supply inventory through UNIFlow and capturing and analyzing more of our instrument QC data to detect problematic trends before they cause failures. We are also looking forward to using UNIFlow's new correlation analysis feature to analyze the total history of a set of successful or failed outcomes to identify resources that contribute more than their share to failures, which will greatly simplify troubleshooting. Besides our lab, other lab teams in our organization have purchased their own UNIFlow systems and are preparing to implement their own processes.