Laboratories consume data, generate data and that data is used in decision making. Whether it’s property data for a new chemical entity; a standard test to determine levels of impurity in a material; or targeted experiments optimising a formulation to satisfy a customer’s specification - data is critical at every point. It may be data such as instrument parameters, or a standard operating procedure required to conduct a test or experiment. It may be instrument-generated data, like a UV/visible spectrum or a melting point.
It may be observation data from the individual conducting the experiment and logged in the form of handwritten notes or dictated audio recordings. It can be a library of static images, digital or physical photographs of samples, films and videos of equipment. Certain Health and Safety data must be present to meet legal requirements. Staff rostering, financial and administrative data are business critical. Data touches every part of the lab.
Efficient labs have efficient data management where the required data (originally in paper form, but now much more often digitally) is logically organised, quick to find and easy to use. Data acquisition is easy, but it must be designed with care so that data location and data application/use can be equally productive. This can be achieved by connecting instruments to stream into data lake repositories, and/or adopting electronic lab notebook software platforms with a variety of “human-laboratory data interaction”. No longer restricted to written or typed information, input and output can now include options for voice, gesture and touch input or the rapidly emerging use of AR/VR/MR in the lab.
It all sounds easy and countless films make it look easy - but achieving this Brave New World is not all it seems.
Paper records have to be digitalised for them to be useful in the lab of the future. The QWERTY keyboard/mouse combination is still the most common input method in the lab. Connection of instruments presents major challenges because of the absence of a standard communication format. There is no equivalent of MIDI (Musical Instrument digital interface) for laboratory instruments. Imagine if every petrol station had different sized fuel nozzles at the pump. Filling up would become a nightmare process!
And lastly, simply digitalising your existing workflows does not necessarily make them, or the lab, better. A poorly optimised paper-chain workflow will be almost as inefficient if it is digitalised “as-is”. The starting point should go back to the basic question: “What do I need to achieve in this workflow?” [There are many thousands of labs which run fully digitalised LIMS/LES/ELN systems, but many of them will admit that there are still areas of their workflow where digitalisation hasn’t happened yet] My interest is in helping “analogue” R&D laboratories to examine their current practice and future need and match a full or partial digital solution to it.
