How to watch new employees use smart kitchen equipment for the first time

Just because a touch screen is larger and has more presets doesn’t automatically mean it’s easier for new employees to use. To identify learning barriers, ask an employee who has completed the necessary security training but has never used the interface to try an explicit task in a secure simulation environment and record where he can't find the entrance, misunderstands prompts, needs help, and how to recover. This is called first-use observation, not blind testing: there is no design to hide objects or brand comparisons, so the word "blind" should not be used to increase the sense of experimentation.

The manufacturer's combi oven selection article focuses on ease of use and describes images and preset functions. [1] The observational methods presented in this article do not validate those claims and were not tested on real workers or specific models. It is suitable for helping stores improve training and operating materials, rather than assessing employee capabilities or creating the conclusion that "everyone learns it once."

Not prompting for the task path does not mean not giving security information.

Start by narrowing down the tasks to be observed. If you just want to know if your employees can find menus, understand icons, and recognize end states, use an approved demo mode, interactive prototype, or paper interface. It cannot be assumed that any equipment has a safe demonstration mode, nor can it be temporarily rewired for testing. Actual power on, heating or food operations should have corresponding authorization, training and control and cannot be mixed into the first round of interface exploration.

Participants need to know in advance which actions are prohibited, how to stop when encountering uncertain situations, who to ask for help, and how to quit at any time. "Not giving task path prompts" only means that the observer does not tell in advance where to press each step. It does not prohibit employees from asking for help. Intervene immediately when danger is about to occur, and record the reasons and intervention actions afterwards; safety priority is not affected by test design.

The task card can write "Find the designated practice program in this simulation interface, check the name and enter the confirmation page, and stop before starting" instead of writing "Press the upper left corner, second line, and green key in sequence." But it also can’t be intentionally vague and leave employees guessing about what is to be accomplished. First, ask someone who is not participating in the formal observation to check the meaning of the questions to avoid counting failures caused by poorly written questions as interface problems.

Also confirm which version you are watching. The device software, interface language, training materials, and task card versions should correspond; after updating midway through the test, the results need to be separated. For those who have difficulty seeing the text clearly or are unfamiliar with the interface language, the actual needs should be recorded and reasonable support provided, and no personal data unrelated to this task should be collected.

Only by breaking a minute into events can you know where to change it

It also took a minute. Some people were reading the warning carefully, while others were choosing repeatedly between two similar names. Keeping only the total elapsed time cannot distinguish between the two cases. The observation record should include the time point, the page on which it is located, visible actions, what the participants said at the time, whether they received help, and the results of the task. It is speculated to be marked separately.

For example, if you see an employee pause, first write down "Staying in the program list, not operating" instead of immediately writing "Can't use it." He might be checking the name of the dish, or he might not know how to get back. From a safe, task-free location, ask how you understood the text and match the explanation to the behavior. The observer's own opinion that the meaning of the button was obvious did not substitute for the participant's understanding.

Whether it can recover after accidental touch is also worthy of separate observation. In the first round, a scenario of "the wrong exercise program was selected but not started" can be given in the prototype and employees can explain how to exit and re-confirm. Do not remove sensors, block ventilation, or create real alarms. Paper drills prove understanding of information and procedures, and cannot be used to conclude that employees have real troubleshooting capabilities.

When recording a request for help, the help content should be retained. Just saying "you can read the instructions" is not the same level of support as directly pointing to the correct button; the task completed after the latter should not be marked as a completely independent success. If the completion rate is counted, the denominator is the valid tasks included in advance, security intervention, task aborts and data defects are listed separately, and the number is not increased by deleting difficult records.

A hypothetical breakpoint: Employees interpret “return” as “cancel”

Assume that in a virtual, non-heating interface, participant A needs to check the exercise program and stop before starting it. He found the wrong selection on the confirmation page, paused for about 20 seconds, and then selected "Return", but was not sure whether the selected task was still retained. The observer did not continue to guide him to start, but ended the simulation and recorded the original words: "I thought I would cancel it when I got back." This scene only shows the recording method, not a real defect of a certain device.

The corresponding record is not "employees respond slowly", but "task status is not easy to confirm after page switching". Material revision can add a status description verified by the supplier, and use actual interface screenshots to clarify where the selected content is displayed; you cannot make up a cancellation function that the product does not have. If the feedback on the product itself is unclear, the problem should be fed back to the person in charge. The operation card cannot permanently cover up important interface flaws.

A safe simulation environment was still used during the retest, and the goal was to see whether the participants could accurately explain whether the current task existed and how to reselect within the allowed range. Instead of asking him to memorize a standard answer, observe whether he can obtain consistent information from the interface and documents. If the observer is still required to point out the status, record this dependence and do not use the "final correct answer" as sufficient evidence that the revision has been successful.

This chain should retain four nodes: visible pauses, participant explanations, specific changes to the material, and new behaviors during retesting. Each node has an object, so that "enhanced training" is not used as a general treatment for all problems. For a pause for which the cause cannot be confirmed, you can continue to observe it and do not need to forcefully characterize it as a design error.

First use, next day retest and revised test answer different questions

First-use observations focus on how people who are new to the interface understand the task. Repeat the test with the same materials the next day. What is observed is whether it can be completed again after the previous experience. The results may be affected by practice, memory and the state of the day. If a new task card is provided at the same time the next day, the materials and experience will be changed at the same time, and faster completion cannot be attributed entirely to one item.

When it is necessary to evaluate material revisions, you can consider arranging other participants with similar task backgrounds and who have not been exposed to the old materials to use the new version, while retaining the comparison conditions; if conditions permit, design a more reasonable grouping or sequence. Small-scale work observations are usually not enough to draw strict causal conclusions. Personnel differences, limited samples, and task scope should be clearly stated. It cannot be automatically called a controlled trial just because new participants are added.

The original participant can also continue to participate in the new version of the test, but his feedback is more suitable to answer whether the specific breakpoint still exists. The record should indicate how many tasks have been completed and what help has been received. Do not rank the third proficiency performance directly with another person's first performance, and do not treat multiple operations by the same person as multiple independent employee samples.

A smooth finish the first time does not prove the same success during rush hours, night shifts, or all menus. When you actually transfer to a position, you still need to complete training and qualification confirmation based on equipment and tasks. Interface observation is part of problem detection and does not replace on-site safety management, nor does it replace food process verification.

Reports deliver problem locations, not employee rankings

The conclusion can be written in a specific and restrained way: a certain version of the task card did not indicate the task status after return, which caused a pause in this simulation; which participants can confirm after the revision, which ones still need help, and what to check next. When there is no real test data, only methods and blank fields are provided, and the efficiency is improved without writing.

For records and videos, the purpose, access scope and storage arrangements should be explained in advance, and only necessary content should be retained. Names do not need to appear in summary reports; the use of numbers does not mean that all images and sounds are rendered unrecognizable. Don't use the collected difficult pieces for performance rankings to avoid employees hiding real doubts in order to perform well.

Valuable first-use observations will make training materials, interface descriptions, and help paths more relevant to actual tasks. What should be retained in the end should be why a certain step is easy to misunderstand, how to revise it, and how to confirm it again, rather than a lack of conditions for "the device is smart, so it is easy to use."

References

• Alto-Shaam: 4 Tips for Choosing an Efficient Combi Oven