Food materials and products have various chemical and physical characteristics and changes and their individual differences and characteristics are large. Therefore, analysis, evaluation, and manipulation of food products are often more difficult than for industrial products. In recent years, with the innovative progress of AI, Big Data, and robotics and mechatronics, recognition and adaptation capabilities for unknown, multiple, ununiform and complex objects have been drastically improving. Research and developments in food engineering effectively utilizing such advanced technologies are conducted, and referred to as ‘new food engineering.’ A Cabinet Office is now leading large-scale projects to cope with food production and consumption problems by employing systems highly integrating cyber space and physical space. Such large-scale projects are on a cutting edge frontier where enterprising researchers develop new food engineering technology. The new food engineering can be regarded as research and development in which machines instead of humans make and eat food to contribute to design, development, and production of foods. Therefore, actual research and development tasks are classified into two groups: ‘making’ and ‘eating.’ Finally, it should be noted that true growth of new food engineering is achieved through experiences where food science researchers and engineering researchers work together, and share failures and successes.

抄録全体を表示

This paper describes soft robotic hands for handling of food materials. Recently, the food industry required automatic handling of food materials to cope with labor shortage and to reduce contagion risk. Despite this requirement, it is still difficult to perform automatic handling of food materials. This difficulty originates from large variances of shapes and dimensions and uncertainty in positioning of food materials. One solution to automate food handling given the large variations in shapes, dimensions, and positioning is to introduce soft materials into robotic hands. Soft materials, acting as a physical interface between robotic hands and food materials, are effective in dealing with these variations. In this paper, we introduce soft robotic hands including wrapping grippers, circular shell grippers, planar shell grippers, binding hands, scooping-binding hands, and needle grippers, which are designed and developed for handling of food materials.

抄録全体を表示

In terms of cost, productivity, and visual appearance, it is difficult for 3D printed food to compete with existing common food products. However, it may become popular for use in special environments such as for artificial cultured meat and space food. In addition, considering the fact that food 3D printers can automatically supply food with tailored flavor, aroma, and hardness to individuals, food 3D printers can be actively used for making nursing care food. We have developed a screw-type, dual-nozzle food 3D printer that can be used to make soft nursing-care food with an external shape reminiscent of actual food. This printer can also create food texture that cannot be produced in actual food material by printing hard and soft materials in three dimensions.

抄録全体を表示

Consumer demand for delicious food is ever increasing. In order to respond accurately and promptly to such demands from consumers, it is essential to rationally improve the quality of food. In such a situation, conventional thinking and empirical knowledge are not enough, and artificial intelligence (AI) is attracting attention as a new tool. Currently, AI is being actively introduced in industrial food production lines for the detection of foreign substances and visual inspection of materials and processed foods. On the other hand, in the fields of food processing and evaluation, there are not so many examples of AI utilization other than those mentioned above. This means that although AI has been very well developed in recent years, there are still challenges. In this paper, we outlined AI and introduced three case studies from the research topics we are working on. First, we explained how to analyze the relationship between food processing conditions and evaluation values obtained by instrument measurement using AI, taking as an example the effect of the structure of gluten protein inside noodles on its texture. Second, we explained how to utilize AI for instrument measurement of foods, taking as an example the development of a bread quality evaluation method based on hollow structures. Third, we explained how to incorporate AI into sensory evaluation, taking as an example the development of a method for measuring the food palatability that human perceive when chewing.

抄録全体を表示

Humans perceive taste, odor, and food texture during mastication. In particular, food texture is an important quality indicator in determining the palatability of food. Hence, food developers require an evaluation method for food texture. Mechanoreceptors in the periodontal ligament contribute to the delicate perception of texture and are roughly divided into two types of response characteristics: fast adaptation, and slow adaptation. To measure and evaluate food texture, a magnetic texture sensor is designed to imitate the characteristics of such mechanoreceptors and the structure of teeth. The magnetic texture sensor mainly consists of a probe, a linear slider, a spring, and a circuit board. The probe is made by a 3D printer, and any shape can be formed on the probe tip for the measurement of various food. The probe also contains a permanent magnet. Both ends of the spring are fixed to the probe and the circuit board. The linear slider limits the oneaxis displacement of the probe, and the bearing balls on the inner surface of the linear slider make the smooth displacement of the probe. The circuit board has two types of sensor elements, magnetoresistive elements, and an inductor. Magnetoresistive sensors and inductors measure the displacement and velocity of the permanent magnets in the probe as changes in magnetic field strength and magnetic flux density, which correspond to the receiving characteristics of slow-adapting and fast-adapting mechanoreceptors, respectively. This paper firstly describes the structure and measurement characteristics of the magnetic texture sensor and shows an example of measuring food. Furthermore, the method of estimating the texture evaluation value from the measured data using Gaussian process regression is described, and the estimation result using commercially available foods is also introduced.

抄録全体を表示

Food texture still relies mainly on sensory tests. During the past decades, texture profile analysis (TPA) has been developed using instrumental measurements called food compression tests using a texture analyzer. Although TPA has been widely applied for a variety of foods, TPA and multi components analysis for characteristic values estimated from TPA do not work well for some foods with complicated food texture. For snacks, such as potato chips and wafers, fracture pattern varies on each piece, and the distribution of estimated values from TPA is widely spread. Machine learning (ML) may have potential to resolve the complication of the food texture analysis. Conventional TPA uses only a few (or up to 10) characteristic values. ML can handle all the data points, such as all the data force values, up to thousands of values from one compression measurement without manual selection of the important characteristic values. ML was applied to discriminate the types of potato chips and/or the other snacks, and it was demonstrated that ML could work for the discrimination of the types of snacks even for foods which are difficult to analyze allowing discrimination of foods based on TPA.

抄録全体を表示

Various kinds of amines are consumed by us in everyday foods. The origin of amines in foods may be as natural ingredients in plants and animals or produced by microorganisms in processing and preservation of foods. Biogenic amines, including histamine, tyramine, putrescine, cadaverine, spermidine and spermine are nonvolatile amines frequently occurring in fish products and fermented foods. For better or for worse, these amines influence our health. It is therefore necessary to know the content of amines in food for risk evaluation and maintenance of human health. In this article, nonvolatile amines in foods and their influence on our health are reviewed.

抄録全体を表示

PC-12 cells were used to study the mechanism of cell death by apoptosis. Crocin showed strong anti-oxidant activity. Crocin decreased the level of ceramide by inhibition of sphindomyerinase activity and increased glutathione concentration resulting in inhibition of apoptosis. Saffron extract and/or crocin ameliorated learning and memory performance inhibited by ethanol in mice. Long term potentiation blocked by ethanol shows recovery by the addition of saffron extract and/or crocin to mouse hippocampus tissues. Mice injected with crocin expressed non-rem sleeping even during an activating period. The clinical trial of saffron extract ameliorated symptoms in Alzheimer’s patients. A Chinese formula prescribed with saffron reversed brain vascular dementia. From these findings saffron and crocin can potentially be applied for treatment and prevention of dementia in patients.

抄録全体を表示

Recent climate change has had a significant impact on marine ecosystems. Seaweed production has become less stable and is declining in many traditional seaweed producing areas along the coast of Japan. Against this background, land-based production of seaweed, which can be systematically and stably produced, is increasingly an attractive alternative production method. The Kochi University Marine Botany Laboratory has established a floating seaweed land-based production method which consists of a ″germling cluster method″ proposed for dense tank culture of seaweed in free-floating form and a multistep tank system. In Muroto City, Kochi Prefecture, for the first time the land-based production of an expensive green laver, Ulva prolifera, was commercialized using nutrient rich seawater pumped from over a 300 m depth. After that, several production facilities were put into operation, in which the cost of the capital investment was kept low by using saline groundwater and changing the material from expensive FRP to lowpriced polycarbonate. Most of the production of U. prolifera in Kochi Prefecture has been subsequently replaced by onshore production. The Kochi method has spread nationwide, and as of 2021, an annual total of 20-30 tons of the green laver has been produced. Research is underway to improve the production system and improve efficiency with the aim of applying it to other types of Ulva such as purple laver and using it for purposes other than human nutrition, such as feed for cultured mollusks and for industrial uses.

抄録全体を表示

The physical properties of starch granules of cereals (barley, rice, wheat, corn), pseudocereal (quinoa), rhizomes (potato, sweet potato, yam, cassava, lotus) and beans (azuki, kidney bean, pea) were compared. The shapes of starch granules were characteristic of the respective plant species. The granule sizes ranged from 1 μm to less than 100 μm. Generally, the distribution patterns of granule sizes showed a single peak, but the distribution patterns of barley and wheat granules showed three peaks. The gelatinization temperature and enthalpy showed a tendency to increase in order from cereals to beans to rhizomes. The water absorption power of the granules increased in the order of azuki, wheat, barley and potato at 20°C. The digestion of raw starch granules by β-amylase resulted in two patterns of degradation, the outer and interior degradation types. The results suggest that the structures of plant starches are diverse and complex.
The structure and functionality of large, medium and small granules from normal and waxy barley were examined. The median size of large, medium and small granules was 18.1, 11.4 and 2.2 μm and 17.2, 9.2 and 2.0 μm in normal and waxy starches, respectively. The amylose content was 25.9% for normal small starch granules, about 4% lower than that of large and medium granules. For waxy starches, the apparent amylose content ranged from 2.9〜3.3% and was similar among the granules. The average chain length (DPn) of amylopectin was 6100〜8900, and similar between the varieties but differing among the granules. The chain length (CL) of amylopectin was 17〜20 glucose residues, being shorter in small granules than in large and medium granules. The relative crystallinity was 20.3 〜23.9% for normal starch granules, and 33.0〜37.1% for waxy starch granules. The transition temperatures and enthalpy changes were 55.6〜71.7°C and 7.4〜8.5 J/g, and 58.9〜77.2°C and 10.1〜12.1 J/g for the normal and waxy starch granules, respectively. The small granules displayed the greatest susceptibility to enzymes and the fastest retrogradation both in normal and waxy starch granules. The difference in functionally among granules of the same variety exceeded the differences between the same granule sizes of different varieties.

抄録全体を表示