New machine paves manner for know-how to ease each day meals, beverage monitoring for seniors, chronically unwell

College of Maine researchers invented a tool that would function a launchpad for brand new know-how to ease each day monitoring of meals and beverage intake for seniors and other folks with continual well being prerequisites.

SipBit, created by way of Nimesha Ranasinghe, an assistant professor of spatial computing, and Chamath Amarasinghe, a Ph.D. scholar of spatial knowledge science and engineering, can pinpoint beverage kind, quantity and sugar content material as soon as submerged in liquid. The machine, educated to acknowledge those characteristics via a sequence of deep studying algorithms, can determine them by way of the usage of electric impedance measurements over a variety of frequencies of a drink.

Electric impedance is the opposition presented to the electrical present—transferring charged debris like electrons— by way of an object. A given set of measurements of {an electrical} impedance correlates with a selected beverage trait, equivalent to whether or not it is tea, espresso or soda.

SipBit applies a recognized electrical pulse throughout 1000’s of frequencies after which measures the impedances throughout the drink, one way known as electric impedance spectroscopy. After gathering more than one units of measurements, SipBit analyzes them to spot the bodily and chemical houses of a beverage, specifically what form of drink it’s and its quantity and sugar content material.

Many seniors and other folks with strict diets, together with those that have most cancers, middle and kidney sicknesses, diabetes and different prerequisites, need to manually file each day what they devour and drink intimately to assist maintain their well being, which Ranasinghe says will also be tedious, time eating and open to miscalculation. He and Amarasighe hope to beef up the know-how in SipBit to spot much more characteristics, equivalent to sodium (i.e., salt), carbohydrates, protein, and spice content material, and file it on units like computer systems and smartphones to relieve their burden. They then hope to broaden cutlery and drinkware which can be provided with SipBit.

“We are ranging from this basis,” Ranasinghe says. “Our subsequent step is to discover other programs and create a wise cup or tumbler the place other folks can mechanically file their calorie consumption intimately and actual time.”

SipBit differs from different know-how designed to spot quite a lot of facets of meals and drinks by way of the usage of electric impedance spectroscopy as an alternative of inspecting footage of what other folks devour and drink prior to and after foods.

Amarasinghe and Ranasinghe say different units have issues correctly distinguishing drinks of the similar colour and figuring out their quantity and different inside attributes. As an example, they are going to battle to resolve whether or not a espresso was once a dismal or medium roast or how a lot is the sugar content material of the drink.

“Via electric impedance, we will be able to acknowledge other beverage sorts and characteristics in keeping with other signatures; for instance, a cup of Pepsi as opposed to a cup of Coca Cola,” Amarasinghe says.

Ranasinghe’s analysis basically makes a speciality of replicating culinary reports by way of simulating tastes, smells and different senses via multisensory augmented fact and different applied sciences. His innovations have earned awards and nationwide acclaim. A few of his in style units come with the Digital Cocktail, or Vocktail, Virtual Lollipop and Digital Lemonade.

Ranasinghe says SipBit builds on his earlier analysis as a result of having sensory know-how that may hit upon and accumulate knowledge on other meals or beverage characteristics will assist invent new units to simulate the ones reports. He says he in the end hopes to broaden a digital area the place other folks can percentage simulated meals and drinks all over the world.

“We will be able to supply new alternatives for human-food interactions,” Ranasinghe says. “One force I’ve is determining how we will be able to upload those unexplored senses—scent and style—into our applied sciences. SipBit is a part of a larger puzzle of long term human-food virtual interactions as a result of if you wish to incorporate new senses, you must invent know-how that may simulate them, sense them and accumulate knowledge on them.”