Um roba´ 'chef' foi treinado para provar alimentos em diferentes esta¡gios do processo de mastigaa§a£o para avaliar se estãosuficientemente temperado.
![](data:image/jpeg;base64,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)
O chef roba´, que já foi treinado para fazer omeletes com base no feedback do provador humano, provou nove variações diferentes de um prato simples de ovos mexidos e tomate em três fases diferentes do processo de mastigação e produziu 'mapas de sabor' dos diferentes pratos . Os pesquisadores descobriram que essa abordagem de “sabor a medida que vocêvai†melhorou significativamente a capacidade do roba´ de avaliar com rapidez e precisão a salinidade do prato em relação a outras tecnologias de degustação eletra´nica. Crédito: Bio-Inspired Robotics Laboratory, Universidade de Cambridge
Um roba´ "chef" foi treinado para provar alimentos em diferentes esta¡gios do processo de mastigação para avaliar se estãosuficientemente temperado.
Trabalhando em colaboração com a fabricante de eletrodomanãsticos Beko, pesquisadores da Universidade de Cambridge treinaram seu roba´ chef para avaliar a salinidade de um prato em diferentes esta¡gios do processo de mastigação , imitando um processo semelhante em humanos.
Seus resultados podem ser aºteis no desenvolvimento de preparação de alimentos automatizada ou semiautoma¡tica, ajudando os robôs a aprender o que égostoso e o que não tem, tornando-os melhores cozinheiros.
Quando mastigamos nossos alimentos, notamos uma mudança na textura e no sabor. Por exemplo, morder um tomate fresco no auge do vera£o liberara¡ sucos e, a medida que mastigamos, liberando saliva e enzimas digestivas, nossa percepção do sabor do tomate mudara¡.
O chef roba´, que já foi treinado para fazer omeletes com base no feedback de provadores humanos, provou nove variações diferentes de um prato simples de ovos mexidos e tomates em três etapas diferentes do processo de mastigação e produziu "mapas gustativos" das diferentes pratos.
Os pesquisadores descobriram que essa abordagem de "sabor a medida que vocêvai" melhorou significativamente a capacidade do roba´ de avaliar com rapidez e precisão a salinidade do prato em relação a outras tecnologias de degustação eletra´nica , que testam apenas uma única amostra homogeneizada. Os resultados são relatados na revista Frontiers in Robotics & AI .
A percepção do paladar éum processo complexo em humanos que evoluiu ao longo de milhões de anos: A aparaªncia, o cheiro, a textura e a temperatura dos alimentos afetam a forma como percebemos o paladar; a saliva produzida durante a mastigação ajuda a transportar compostos químicos dos alimentos para os receptores gustativos principalmente na langua; e os sinais dos receptores gustativos são passados ​​para o cérebro. Uma vez que nossos cérebros estãocientes do sabor, decidimos se gostamos da comida ou não.
O sabor também éaltamente individual: algumas pessoas adoram comida picante, enquanto outras gostam de doces. Um bom cozinheiro , seja amador ou profissional, depende de seu paladar e pode equilibrar os vários sabores de um prato para fazer um produto final bem arredondado.
Â
“A maioria dos cozinheiros caseiros estara¡ familiarizada com o conceito de degustação a medida que vocêavana§a osverificando um prato durante o processo de cozimento para verificar se o equilabrio de sabores estãocorretoâ€, disse Grzegorz Sochacki, do Departamento de Engenharia de Cambridge, o primeiro autor do artigo. "Se os robôs forem usados ​​para certos aspectos da preparação de alimentos , éimportante que eles sejam capazes de 'saborear' o que estãocozinhando."
"Quando provamos, o processo de mastigação também fornece feedback contanuo ao nosso cérebro", disse o coautor Dr. Arsen Abdulali, também do Departamento de Engenharia. "Os manãtodos atuais de teste eletra´nico tiram apenas um aºnico instanta¢neo de uma amostra homogeneizada, por isso queraamos replicar um processo mais realista de mastigação e degustação em um sistema roba³tico, o que deve resultar em um produto final mais saboroso".
Os pesquisadores são membros do Laborata³rio de Roba³tica Bio-Inspirada de Cambridge, dirigido pelo professor Fumiya Iida, do Departamento de Engenharia, que se concentra no treinamento de robôs para realizar os chamados problemas de último metro que os humanos acham fa¡ceis, mas os robôs acham difacil. Cozinhar éuma dessas tarefas: testes anteriores com seu "chef" roba´ produziram uma omelete aceita¡vel usando o feedback de provadores humanos.
“Precisa¡vamos de algo barato, pequeno e rápido para adicionar ao nosso roba´ para que ele pudesse fazer a degustação: precisava ser barato o suficiente para usar em uma cozinha, pequeno o suficiente para um roba´ e rápido o suficiente para usar enquanto cozinhamosâ€, disse Sochacki. .
Para imitar o processo humano de mastigar e saborear em seu chef roba´, os pesquisadores anexaram uma sonda de conduta¢ncia, que atua como um sensor de salinidade, a um braa§o roba³tico. Preparavam ovos mexidos e tomates, variando o número de tomates e a quantidade de sal em cada prato.
Usando a sonda, o roba´ "provou" os pratos em forma de grade, retornando uma leitura em apenas alguns segundos.
Para imitar a mudança na textura causada pela mastigação, a equipe colocou a mistura de ovos em um liquidificador e fez o roba´ testar o prato novamente. As diferentes leituras em diferentes pontos de “mastigação†produziram mapas gustativos de cada prato.
Seus resultados mostraram uma melhoria significativa na capacidade dos robôs de avaliar a salinidade em relação a outros manãtodos de degustação eletra´nica, que geralmente consomem tempo e fornecem apenas uma única leitura.
Embora sua técnica seja uma prova de conceito, os pesquisadores dizem que, ao imitar os processos humanos de mastigação e degustação, os robôs acabara£o sendo capazes de produzir alimentos que os humanos apreciara£o e podera£o ser ajustados de acordo com os gostos individuais.
"Quando um roba´ estãoaprendendo a cozinhar, como qualquer outro cozinheiro, ele precisa de indicações de quanto bem foi", disse Abdulali. "Queremos que os robôs entendam o conceito de sabor, o que os tornara¡ melhores cozinheiros. Em nosso experimento, o roba´ pode 'ver' a diferença na comida a medida que émastigada, o que melhora sua capacidade de saborear."
"A Beko tem a visão de trazer robôs para o ambiente domanãstico que sejam seguros e fa¡ceis de usar", disse o Dr. Muhammad W. Chughtai, Cientista Saªnior da Beko plc. "Acreditamos que o desenvolvimento de chefs robóticos desempenhara¡ um papel importante em famalias ocupadas e residaªncias assistidas no futuro. Esse resultado éum salto na cozinha roba³tica e, usando algoritmos de aprendizado profundo e de ma¡quina, a mastigação ajudara¡ os chefs robóticos a se ajustarem. gosto por diferentes pratos e usuários."
No futuro, os pesquisadores estãoprocurando melhorar o chef roba´ para que ele possa provar diferentes tipos de alimentos e melhorar as capacidades de detecção para que possa provar alimentos doces ou gordurosos, por exemplo.