FEATURE SPECTROSCOPY
Better, smaller, cheaper: why spectroscopy is spreading
From agriculture to AI, spectroscopy improvements are driving the technology into new areas. Experts shared new applications, latest developments and looming challenges with Matthew Dale during a recent Electro Optics webinar
M
odern spectroscopy solutions offer continually increasing sensitivity, resolution, accuracy, speed, integrability, affordability
and user friendliness, all in ever-smaller physical footprints. This has led to the technology becoming adopted across a wide range of industries and applications, both in the lab and out in the field. Such was shared in a webinar recently
hosted by Electro Optics that covered the latest applications, developments and challenges of spectroscopy. According to speaker Nazanin Shafiee, Director of Strategy & Business Development at Avantes, the future continues to look bright. Avantes expects an increase in the uptake of spectroscopy technologies across the industries it serves, including semiconductor, medical, pharmaceuticals, biotech, food & agriculture and environmental monitoring. Nor is this confined to mature companies that have long considered spectroscopy a crucial and well-established part of their operations. Start-ups and scale-ups are also adopting
‘What we can now do is combine AI with spectroscopy to develop intelligent data processing algorithms’
14 Electro Optics May 2023
the technology in new and innovative ways. As modern spectrometers become smaller and more portable end users can more easily integrate the technology into handheld devices and tools for new applications in the field.
Consistent spectroscopic data is key Shafiee highlighted improvements in data analysis and interpretation using artificial intelligence (AI) as important developments. “This is quite significant, because with
the emergence of AI, what we can now do is combine it with spectroscopy to develop algorithms using machine/deep learning,” she explained. “These algorithms are developed and trained using collected spectroscopic data to recognise patterns on key indicative metrics that our customers are looking for, or find new metrics that they may not have previously looked for.” In developing such algorithms, it’s
important that the training data is consistent and reliable – not only when obtained from one spectrometer over a period of time, but potentially from a number of different systems in the same setting. Shafiee said the data must have “as reduced standard deviation as possible” to ensure it is of high enough quality to be used for training algorithms. To facilitate this, Avantes has upgraded its production capabilities – manufacturing process controls, sub-assembly levels and automated
@electrooptics |
www.electrooptics.com
Page 1 |
Page 2 |
Page 3 |
Page 4 |
Page 5 |
Page 6 |
Page 7 |
Page 8 |
Page 9 |
Page 10 |
Page 11 |
Page 12 |
Page 13 |
Page 14 |
Page 15 |
Page 16 |
Page 17 |
Page 18 |
Page 19 |
Page 20 |
Page 21 |
Page 22 |
Page 23 |
Page 24 |
Page 25 |
Page 26 |
Page 27 |
Page 28 |
Page 29 |
Page 30 |
Page 31 |
Page 32 |
Page 33 |
Page 34 |
Page 35 |
Page 36 |
Page 37 |
Page 38 |
Page 39 |
Page 40 |
Page 41 |
Page 42
