It is safe to say that automation has an established position in several industries to varying degrees. Saving time or lowering costs, rearranging resources, getting rid of tedious or repetitive tasks or minimizing process error sources are probably some of the most usual advantages connected to automating processes. When it comes to our specific and sensitive E-TRF method, minimizing all environmental or process factors that could end up affecting the results is extremely important.
Our Application Scientist Matias wrote down his thoughts about processing large sample quantities, process errors, ensuring high-quality results, and integrating automated systems to our methods, and how we can utilize it in our laboratory.
Processing a larger number of samples efficiently
During the different validation phases of our development work at Aqsens Health, the analyzed sample amounts are somewhere between hundreds and some thousands. However, when developing screening tests accessible to the population level, as time goes on, the amount of performed tests is only going to increase. This means that a method like our E-TRF must be low in costs and scalable – two traits very often connected with an automated process.
With the increasing amount of performed measurements, the likelihood of human error also usually tends to increase due to the repetitive nature of the performed tasks. Taking this burden off the employees and allowing them to concentrate on other aspects of the development process not only increases the efficiency of the development workflow, but also results in faster and better measurements results in terms of reproducibility.
From a business perspective, automated systems could be seen as an investment that in the long run pays for itself. When talking about scalability, there is certainly no question about if it is easier to recruit more people to work on the same tasks or to just upgrade an existing automated system, if the end result is the same. Arguably, even in smaller companies where personnel or capital are limited, it is vital to use them where they end up mattering the most.
Minimizing errors and ensuring high-quality
Usually, errors are divided up into random, systematic or personal errors. In short, random errors are fluctuations in the results and they can be analyzed statistically when the amount of measurements is increased. Systematic errors are in turn errors, which repeatedly come up in every measurement and alter the results consistently in the same way.
Human error comes into account when analyzing the behaviour of the experimenter in the measurement process; carelessness, technique, experience and other human aspects are also part of the result as a whole. Making mistakes tends to be a human trait and especially this part of the total error can be minimized through automation of the measurement process. Automated systems rarely make mistakes and always pipette in the same manner, without small variations that can occur when the work is done by hand. This ensures that the quality standards remain high and consistent across thousands of sample measurements.
Automation in our research and development process
In our laboratory, we ended up acquiring a Fluent – workstation manufactured by the Swiss company Tecan Group Ltd. Out of the several alternatives available, some of the properties of the Fluent – system were suitable for us. An open and modular system allows us to adapt our measurement process onto the platform. The open system also makes modifications later on in the development process pipeline possible. Integrating some parts of our already existing methods straight onto the platform provides a smooth transition from a manual to an automated process with minimal hassle.
For our work, minimizing measurement errors and later on reducing the process time are the two aspects of automation that I would consider the most important. Minimizing the effect of environmental or other process factors on the measurements ensures the reliability and accuracy of the results. Reducing process time and amount of repetitive work through automation enables efficient and large-scale operations, without compromising the quality of work.
As an example, the routine work when analyzing a sample set of around 300 samples manually would take us around 22 hours, three days of work in the laboratory, even when the method is considered fast and simple. Based on early estimates and established automated methods, the same amount of samples could be analyzed and measured in five to six hours using the automated workstation. Who wouldn’t want to get three days of work done in just one day?
When analyzing samples for information about the health of a person, it is extremely important to get the job done right the first time. If there is an indication of a disease, but it ends up being a falsely positive result or a falsely negative one, the implications could be severe. If treatment of a patient is initiated on an uncertain basis, it can end up not only wasting important resources, but also might be completely unnecessary or not targeted towards the core issue. Accuracy and reproducibility in our measurements contributes towards more certainty in the statements based on the available data from the samples. Minimizing the measurement errors lies at the heart of producing accurate results.
Our current work focuses on ensuring that the automated process ends up meeting the requirements we have set. In practice this means setting up and performing different tests in combination with our manual measurement process, and ensuring that the automated version performs at least as well or even better than our manual work in the laboratory. The quality of results is something that we need to be certain of and can make no compromises in – the other improvements of automation come in second.
Eventually when the automation process is complete and the entire measurement platform is established, extending automation to more laboratory functions is probably the natural follow-up step. The ultimate goal could be that the entire process from sample management to delivering results to healthcare professionals and patients would be automated, where applicable.
Personally, I am looking forward to having an independently functioning system that makes everyday life at laboratories easier, and a complete package which can be distributed to other research groups and operators so that eventually the work that has been done ends up improving the quality of life at the population level.
Aqsens Health Oy
Aqsens Health was recently featured in the Tecan Journal, where our time was interviewed about the Fluent workstation and why we chose to add it to our laboratory. Read the article on the Tecan Journal’s website: https://www.tecan.com/tecan-journal/non-invasive-screening-tests-for-preventive-healthcare