Cooking is an essential part of most people’s day-to-day lives. It is a dynamic activity that involves multitasking, timing of tasks, and flexibility. However, the available instruction methods for cooks who wish to follow recipes are quite rigid and linear. This project aims to create a design solution that provides instructions in a way that is in better accordance with natural cooking behaviours.

The scope of the project is to deliver a conceptual instruction-based cooking tool that utilizes Artificial Intelligence. While designing the solution, speculative design methods are considered in order to design for the near future. It is designed for a future where tasks in recipe instructions can be used more flexibly in digital products and Human Activity Recognition can accurately recognize complex cooking actions with mobile devices. The design solution has been created for a company called Verdify. Verdify is a food tech company with products that aim to allow users to achieve fully personalized nutrition. Swapmeals is their recipe platform that allows users to swap ingredients to fit their dietary needs while adjusting instructions accordingly and maintaining flavour profiles. The design solution aims to ultimately live on this recipe platform. The project is executed with a User-centered Design approach which involves end-users in the design process in order for them to have an impact on their final product.

The users were observed while cooking with instructions in order to understand the pain points that should be designed for. The most prevalent and significant themes that emerged from the study were that users experience cognitive overload while cooking and often experience uncertainty in the cooking process.

Academic design solutions and commercial products for hands-free instruction-based cooking were reviewed to understand which issues, discovered in the context research, have not been designed for. This gap is used to specify design requirements. The requirements specify that the design solution should allow home-cooks to use recipe instructions in a way that is hands-free, non-linear, informative, and that minimally interrupts the cooking flow along with reducing their cognitive load and the associated negative emotions. The solution is also required to allow the user to mainly focus on the task at hand; the device should manage the remaining responsibilities. Lastly, the research discovered that the design solution should create the experience of cooking a new recipe similarly to how one would cook a familiar recipe.

Solutions were explored that meet the specified requirements. The exploration led to 2 contrasting concepts that aim to improve instruction-based cooking in a familiar textual manner and an innovative experiential manner. These 2 concepts were tested with users in order to compare the newly defined interaction methods and understand which interactions are desired in the various moments of the cooking process. The results generated an iterated list of requirements to be applied to the final design.

Lastly, a conceptual instruction-based cooking design solution was introduced; a user interface that utilizes Human Activity Recognition for user-to-product communication. The tool adopts the cognitive load from its users by managing timing and their progression while adjusting to the user’s autonomous adaptations to the recipe. The design solution is an interactive, hands-free, and informative recipe instruction solution. It allows home-cooks to let go of keeping an overview of the cooking process and instead solely focus on the task at hand, allowing cooking to be a worry-free and enjoyable activity for even the tensest home-cooks.