The Raspberry Pi-Powered π-LAB Brings Tech to Bear on the Thorny Problem of Olive Oil Quality

Born of an argument regarding homemade olive oil quality, this 3D-printed lab equipment includes built-in AI smarts.

Scientists Georgios Violakis and Nikolaos Vardakis have landed on Kickstarter with a kit to turn a Raspberry Pi into a home chemistry lab: the aptly-named π-LAB.

"π-LAB was born out of pure scientific curiosity," Violakis explains. "Two years ago, a group of friends (scientists and engineers) debated over the quality of the olive oil they produced and — naturally — everyone had a strong argument on why their olive oil was better."

"The question that came up was the following: Could there be a relatively low-cost method to determine — quantitatively — the quality of products we produce or use daily in our kitchen (olive oil, coffee, wine, honey, etc.)?"

Doing so in a lab is easy, through devices like Raman spectrometers or polymerase chain reactors — not commonly found in a home kitchen. The π-LAB is Violakis' answer: A low-cost Raspberry Pi 3 Model A+-powered device that uses computer vision to analyze liquids — and which can, finally, resolve the argument over whose oils are superior.

"It uses a novel optical-filter-on-CMOS spectrometer in combination with a laser, broadband light sources, a camera and the processing power offered by the Raspberry Pi single board computer — and a bit of artificial intelligence magic — in order to perform absorption, fluorescence, and computer vision analyzes of liquid food samples (and any liquid, actually)," Violakis writes.

"The idea is the combination of several analytical techniques and smart post-processing to outweigh the lack of resolution inherent to low-cost equipment, in order to determine the quality of liquid foods with the best possible accuracy."

The open source project has reached the level of maturity where it can analyze the quality of olive oil, using a built-in database pre-populated with samples, in a single push, offers graphical comparatives, and provides the ability to capture similar databases for any other liquid type — from wine to fruit juice.

The system can also use an on-board artificial intelligence to spot different qualities, provide a real-time view of the liquid under laser or white light illumination, provide color and laser penetration differences, and export data for post-processing.

"It should be noted that π-LAB is by no means a substitute of high performance analytical equipment costing tens or hundreds of thousands of dollars (Raman/FTIR spectroscopy, NMR, PCR, etc.) and as such no-one should expect it can perform similar to any of the above devices," Violakis warns. "π-LAB has proven in our tests to be an excellent screening tool, but in no way can it replace a rigorous laboratory analysis."

The project is now funding on Kickstarter, with physical rewards starting at €339 (around $400) for early-bird backers. Hardware is expected to start shipping in January next year.

Gareth Halfacree
Freelance journalist, technical author, hacker, tinkerer, erstwhile sysadmin. For hire:
Latest articles
Sponsored articles
Related articles
Latest articles
Read more
Related articles