Evaluation

The evaluation of the AR Cooking prototype aimed to assess its effectiveness, efficiency, and user satisfaction. The primary goals were to determine how well the app guided users in following recipes and completing cooking tasks, measure the time and errors associated with using the app compared to traditional methods, and understand participants’ overall impressions of the app’s usability and features.

To test the prototype, two key tasks were designed. The first task involved preparing ingredients with AR assistance. Participants opened the AR Cooking app, selected a recipe, scanned ingredients using their camera, and followed AR overlays to chop, measure, or prepare each item. They also used voice commands like "Next step" to proceed through instructions. The second task involved cooking the recipe itself. Participants followed AR instructions to perform cooking techniques such as boiling or mixing ingredients. They also tested the app’s substitution feature for unavailable ingredients and completed the recipe by confirming the final step in the app.

The evaluation included 25 participants between the ages of 18 and 45 with beginner to intermediate cooking skills. Participants were recruited through cooking classes, social media, and local community groups to ensure a diverse sample. All participants had basic familiarity with smartphones, but no prior experience with AR apps was required. The evaluation measured effectiveness by examining the success rate and critical errors during tasks, efficiency by assessing task completion time and error rate, and satisfaction by surveying ease of use, engagement, and likelihood to recommend the app. The study was conducted in person in a controlled kitchen environment, with researchers observing and recording interactions while providing minimal intervention. Data was collected using a think-aloud protocol, post-task surveys, and direct observation.

The results demonstrated high levels of effectiveness, with 88% of participants successfully completing recipes without external assistance. Critical errors were minimal, averaging only one error per user, often related to the need to revisit a step. Participants praised the clarity of AR overlays and step-by-step guidance, stating that the app reduced confusion compared to written recipes or video tutorials. Regarding efficiency, participants completed recipes 30% faster than traditional methods, as AR Cooking streamlined multitasking by eliminating the need to pause and refer to external resources. Minor errors, such as misinterpreting instructions or failing to recognize voice commands, averaged two per recipe and were primarily due to participants’ unfamiliarity with AR technology, which improved with usage.

In terms of satisfaction, the app received overwhelmingly positive feedback. On a scale of 1–5, 88% of participants rated the app a 5 for usability, citing its intuitive interface and clear visual cues. Participants reported high levels of engagement, with comments highlighting the app's ability to make cooking stress-free and enjoyable, akin to having a personal chef. Additionally, 88% of participants stated they would recommend the app to others, while one user suggested improving ingredient recognition for niche items.

Overall, the evaluation revealed that AR Cooking effectively assists users in completing recipes with minimal errors, improves efficiency by reducing time spent on tasks, and provides a highly satisfying user experience. Future improvements should focus on enhancing ingredient recognition for less common items, expanding the recipe database to include more culturally diverse and advanced options, and refining the voice command functionality for greater accuracy. These changes will further optimize the app’s usability and broaden its appeal to a wider audience.