Many labs have taken steps forward in leveraging modern technology to facilitate scientists’ ability to capture their notes, measurements, and observations while performing research. It’s not uncommon nowadays to see tablets, shared workstations, and even augmented reality tools in the laboratory environment.
An emerging trend over the past few years has been to deploy a virtual assistant, most commonly through voice. Here are 3 ways that virtual assistants can help with closing data integrity gaps
Virtual Assistants are often faster and more reliable than manual note-taking. Many labs are still attached to their paper notebooks and other means of capturing notes through pen (masking tape, gloves, etc). Yet due to advancements made in Natural Language Processing tools, data capture through voice has demonstrated that the scientist can record data more quickly and with fewer errors than if they were to do so by pen-and-paper, or even keyboard entry. One customer believes that for every 30 weights recorded by their lab techs, they experience an error rate of 8% when using pen-and-paper. Using a virtual assistant, that error rate drops virtually to nil. Moreover, with LabVoice in particular, we have the ability to send notes via email for users to copy into their ELN or LIMS at the end of the day. Some users have elected to build integrations between our platform and their LIMS so this process happens automatically, saving even more time.
For the integrated lab environment, Virtual Assistants provide more context than what the instruments alone can share. Instrument software - although much better than where it previously had been - still has a ways to go when it comes to its ability to provide context around the data. Even some of the more advanced instrument software cannot provide a complete picture without commentary from the scientist. For example, in animal facilities, the person recording qualitative observations and other metadata may not be captured. Virtual assistants offer an easy-to-deploy setup for scientists to add this context in an integrated lab environment. Even with a complete, standardized data packet coming from the instruments, the ability for data producers to add their observations simply by speaking will help the data consumer better understand what he or she is looking at, also improving collaboration and productivity, as they need to spend less time together reviewing the basics.
Virtual Assistants introduce processes where the process does not exist. Take a standard LCMS sample prep workflow. For many labs, if you were to have 5 techs performing the sample preparation, there could be 5 different outcomes, impacting the quality of the experiments and the data captured in that process. It’s not uncommon for any lab process to have multiple manual steps without defined documentation methods and for there to be a general understanding of the process at a high-level, but great variance in single-step details, increasing the likelihood of error propagation. This all impacts the end result. With a modern virtual assistant, users can create what we call at LabVoice “LabFlows” or virtual-assisted lab processes. LabFlows can be as complex as multi-day SOPs or method execution, but in many cases, they’re much simpler tasks, like replacing a checklist or a more routine process like sample preparation. Virtual assistants ensure consistency in process execution, improving data and experiment quality (and in many cases, productivity!).
Interested in learning more about our platform, or the virtual assistant integrations we’ve built? Request to speak with LabVoice or for a copy of our case studies here: info@labvoice.ai