Computer Vision-Assisted Robotized Sampling of Volatile Organic Compounds
Ching-Chi Chan, Noor Hidayat Abu Bakar, Chamarthi Maheswar Raju, Pawel L. Urban
- Year
- 2024
- Citations
- 1
Abstract
In conventional chemical analysis, samples are homogenized, extracted, purified, and injected into an analytical instrument manually or with a certain degree of automation. Such complex methods can provide superior performance in terms of sensitivity or selectivity. However, in some cases, it would be advantageous to possess a method that circumvents those preparatory steps, which require much attention. Here, we present a facile analytical approach to sampling volatile organic compounds (VOCs). Solid specimens emitting VOCs can be dropped onto the drop-off zone at a random position without any special alignment. A computer vision system recognizes specimen position, and a robotic arm places a sampling probe in the proximity of the specimen. The probe aspirates the VOCs─emitted by the specimen─with the aid of a suction force. A portion of the gaseous extract is transferred to the tritium-based ion source of a drift-tube ion mobility spectrometer. The ion mobility spectrum is immediately displayed in the customized graphical user interface (GUI). The sampling system also features a function for flushing extract ducts with hot nitrogen gas. Multiple specimens can be dropped for analysis at the same time. In one embodiment, the system can distinguish fresh meat from spoiled meat. When two meat specimens are placed on the drop-off zone, they are immediately sampled by the robotic arm, analyzed, and labeled on the digital image displayed on the GUI. Thus, the developed autosampling platform provides a hassle-free way of qualitative or semiquantitative analysis of raw specimens.
Keywords
Related papers
Artificial intelligence: a modern approach
1995
Are we ready for autonomous driving? The KITTI vision benchmark suite
Andreas Geiger, P Lenz, R. Urtasun
2012
TensorFlow: Large-Scale Machine Learning on Heterogeneous Distributed Systems
Martı́n Abadi, Ashish Agarwal, Paul Barham +17 more
2016
Vision meets robotics: The KITTI dataset
Andreas Geiger, Philip Lenz, Christoph Stiller +1 more
2013