HOME

News

August 1, 2006

Introducing a Personal Robot with a Sense of Taste

- Identifies food, including wine -

1. Background

NEC System Technologies worked with Mie University to propose a project for the "21st Century Robot Challenge Program - Next Generation Robot: Practical Application Project (Prototype Development Support Business)" led by the New Energy and Industrial Development Organization (NEDO). Entitled, "A Partner Robot with a Sense of Taste", NEC System Technologies and Mie University's development idea was accepted by NEDO in February 2004. NEC System Technologies worked with the Laboratory of Bio information & Food Engineering (BIFE), led by Professor Hashimoto of the Department of Bio resources, Mie University, which is the most advanced research group in the field of food analysis using infrared spectroscopic technology.

NEC System Technologies and Mie University exhibited their achievement in June at EXPO2005 Aichi, Japan as an update to their technology under development. In the exhibition, the robot identified food without opening the package for several types of cheese (Edam, Gouda, Camembert, etc.), meat products (ham, bacon), and bread (pain de mie, baguette, croissant,); it also gave advice regarding both the food and related health issues. A press release was issued along with this event. (June 9, 2005- Development of a Personal Robot with a Sense of Taste)

After the exhibit at EXPO 2005 Aichi, NEC System Technologies and Mie University set a more challenging goal for the robot - to identify different types of wine and the robot's ability to taste the differences in food. Partial results from this effort were reported on July 21, 2006 at a meeting hosted by NEDO where they were announcing the results of the "21st Century Robot Challenge Program: Next Generation Robot: Practical Application Project". To see the announcement for this event held at the Shinagawa Prince Hotel, please view the site below.

2. Food Tasting Robot and the Mechanism for its Tasting Capability

The robot is capable of examining the naming a food, listing the food's ingredients and tasting the food. The robot then gives advice on the food and potential health-related issues. Tasting by the robot involves the analysis of food components. Analysis includes: estimating the major components such as sugar and fat, found in the food, estimating the presence of food components, and estimating the quantities of the identified components. The tasting capability of the robot is not the same as that of a human. The human sense of taste is synthesized by taste cells on the tongue which senses sweetness, sourness, saltiness, bitterness, and tastiness. Humans do not conduct an analysis of a food's components. We may be able to say that the creation of this robot is somewhat comparable to imitating birds in order to design airplanes.

The tasting robot uses infrared spectroscopic technology to analyze the components of a food. It transmits infrared rays at the food and measures the degree of absorption of certain wavelengths (absorbance spectrum). The absorbance spectrum gives out different wave shapes for different foods. We can think of this wave shape as a "food's fingerprint." We know that when a certain molecule is present in a food, a ray of a specific wavelength is absorbed. As a result, the robot can estimate the major components, such as sugar and fat, and the quantities of these components present in a food. The robot is pre-loaded with food information. The robot has already been introduced to many types of food, and it has cataloged the absorption spectra of various foods (the foods' fingerprints), one by one. When a new food is introduced to the robot, it will compare that food's absorption spectrum against the ones the robot has already cataloged and determine which ones are comparable to the food that has just been introduced. Also, the robot looks at the absorption rate for certain wavelengths and thus estimates the amount of each component contained in the food.

3. Point of Improvement: From Tasting Robot to Sommelier Robot

In terms of absorbance spectra, differences among different types of wine are strikingly smaller than those among other types of foods. See diagrams 1 and 2 for details.


Diagram 1	Diagram 2

Diagram 1: Absorbance spectra for different types of cheese
Diagram 2: Absorbance spectra for different types of wine

For this reason it is more challenging for the robot to discriminate between different types of wine than it is to differentiate between other types of food. In order for the robot to be able to achieve the ability to differentiate wine we made several improvements. Also, since the intention was to create a sommelier robot that can taste wine, we also included a feature that selects the wine suited to a customer's taste by carrying on a dialogue with him/her. The sommelier robot's new features for differentiating wine and selecting wine include the following:.

(1) Improvement of the discrimination algorithm
Discrimination is conducted by focusing on the automatic extracted point of maximum difference between the spectra of the wines being tasted.

(2) Refinement of the infrared sensor
Improvements were made in the resolution of the infrared sensor that conducts the spectra analysis. The sensor now has a wider spectrum that covers the range from near-infrared rays to the mid-infrared spectrum. As a result, the discrimination of more types of wine is possible. However, the improvements resulted in a larger sensor which can not be loaded directly onto the robot and must now touch the surface of the wine. The sensor has to be cleaned for each determination. The team continues to refine the infrared sensor in order to resolve this issue.

(3) New functionality for wine selection
The robot is now equipped with the functionality of asking questions to determine a customer's wine preferences. Like a human sommelier, the robot asks a minimum number of questions in order to narrow down the choices among the best wines.

Diagram 3 shows the second-generation tasting robot (sommelier robot), that was recently developed.


Diagram 3: The second-generation tasting robot that was recently developed

4. Future outlook

The use of information technology that imitates the five human senses of sight, hearing, touch, smell, and taste is important in the advancement of a ubiquitous society. NEC System Technologies and its partners will continue to lead the industry in robotic development that supports the prevention of lifestyle-related diseases and help in the treatments for those diseases through dietary therapy and elemental technology.