If you’re not software engineer or data scientist, setting up servers, implementing machine learning and data mining to your work in a laboratory, probably isn’t your forte. Even should you possess these skills, it’s most likely not something you have time for when the next breakthrough could be lurking in the details of your years of hard work!
Our work begins with the person working in the laboratory. We know everybody wants to do good work but everybody makes mistakes. Usually we want to avoid these mistakes, but if that mistake leads to a novel finding we definitely want to know what it was! Recording every action that we take in a lab requires a lot from a computer, but we aim to resolve this by combining machine vision and object recognition with various markers and optimizing the data that’s being analyzed. These things are not science fiction any more.
We are building mathematical models of our experiments and this is our key to unlocking the power of artificial intelligence and letting the machine ”understand” what the person is working on, what is the goal and how we want to get there and what were the results. Developing these data structures for this interaction between us and the machine isn’t all that far fetched.
Following a protocol to the point is how you get those error bars smaller and your quality control the best. Unclear instructions and insufficient documentation are perquisites for failure and this is what we want to erase from your laboratory!