Product Experience In collabration with Philips 2019
Saeco Xelsis is a fully-automatic coffee machine, Philips assigned us a brief of improving the UX of the espresso machine and evaluation of the design. This project was done in the course of Usability testing and user eXperience Assessment in Design (UXAD).
This project went through 2 phases with designs and validation in each of them. Started with the analysis of the current product experience, we ideated 3 concepts to engage users in the design cycle and to find out insights for narrowing own the design space. The converged concept combined insights from test results of previous design concepts. It was later evaluated by testing with users, and future suggestions were proposed in the very end.
Philips Saeco Xelsis is a fully automatic espresso coffee machine that offers a wide variety of coffee. In the first stage, we experienced it, observed how people interact with this product, decomposed it and mapped out possible directions to improve the user experience.
Key findings for improvements
Through the intial empirical studies and research (Read more), we spoted some design flaws and opportunies of contributing to a smooth and satisfying user experience.
While making coffee...
Too many steps before brewing coffee
The machine allows users to do 4 to 5 different customisation. However, the user needs to go through at least 6 to 7 steps before they can get their coffee if they want to do all the customizations.
The profile function is not welcomed
There is a profile function that allows the user to make coffee within 3 steps. However, as the profile section is separated from the customization process and the users are easily confused by this function. Thus, they are not willing to use the profile function.
When maintenance is required...
The maintenance requirements pop up in an unexpected moment
For example, adding water is required after the machine has already started grinding the beans. It annoys the users by breaking their expectation of the coffee.
The design goal is to improve the user experience by making the coffee customization process as well as the maintenance requirements easier to control.
For coffee making Build a better link between the customization process and profile feature.
For machine maitenance Provide beforehand maintenance information.
After the exploration of 3 concepts and tests with peers (Read more), we took away highlights from each of the 3 and have them blended into one new concept that accommodates the majority of our design within a full-color touch screen and a few LEDs embeded in embeded in the maintenance parts.
Highlight 01 - Coffee ID
The Coffee ID is designed to give quick access to the personal preference (profile) of that drink from the drink page. It is designed to be non-intrusive, so it does not limit the coffee-making experience.
The profile panel and customization panel are both placed on the right side of the screen with the same visual style to show the connection between both functions. Each panel can be selected individually. Each panel is displayed in the collapsed view where the selected profile (or default) and an overview of drink settings are presented. They can be expanded to access profiles as a shortcut and coffee customization where coffee settings and profiles are bonded to each other.
The design enables multiple ways to access the profiles and it is closely connected to customization because they are inseparable and in the home context profile is a more frequent feature for lazy coffee lovers to interact with than customization.
Highlight 02 - Maintenance prediction
Maintenance forecast consists of two parts, the maintenance indication bar and the maintenance information page. It is designed to be the main place for monitoring and managing the frequent maintenance of the machine(e.g. Filling water/beans and emptying the coffee grounds tank), aimed at eliminating the hassle caused by maintenance and giving more control over maintenance to users.
The maintenance indication bar
The maintenance indication bar represents the maintenance status so that users can quickly make a decision, whether to do maintenance or not, by showing the number of coffees that can be made with the ingredients left in the machine. It only shows up when the available ingredients can support three drinks or less. It is intended to work as a mild reminder when maintenance is almost required.
The maintenance information page
By tapping on the maintenance indication bar, the maintenance information page can be accessed. This page will also pop-up when maintenance is being done (e.g. water tank is taken out) and when there are not enough ingredients to make the desired drink.
In the maintenance information page, an overview of the amount of water, coffee beans and grounds capacity are displayed and converted to the number of drinks (Americano) that can be made with the available ingredients. By displaying all the information together, users can decide to do other maintenance.
The explanation page of the maintenance information
In the explanation page of the maintenance information, it explains how the machine counts the available cups of coffee. Considered that different kinds of coffee require different amount of ingredients, the ingredients for one cup of Americano (which cost most ingredients) is used as one unit of coffee to guarantee that the user can make whatever what kind of coffee.
Light indication of maintenance
A test plan was made to identify issues in the redesign with regard to interaction and user experience. Further unforeseen problems are taken into account for the final redesign.
Test the highlight features
- Find out whether the users can make their customised coffee easier and more fluently.
- Find out if the maintenance forecast feature makes the maintenance notification more pleasant.
- Find out if the lighting icons help the users during the maintenance period
Evaluate the overall usability
- Find out whether people understand the information, icons and interactions provided by the machine.
- Find out if the aesthetics style of the machine fits the market position of product ( high-end product series)
Task 1 Explore the functionalities of the machine
The goal of the task: Let the user explore the different sections of the machine to get accustomed to the machine.
Task 2 Make an espresso with customised taste
The goal of the task: Let the participant prepare an espresso for themselves and customise the taste. And see if people are willing to use the profile feature to save the drink, and if the indication of how to save a profile is clear.
Task 3 Make two cappuccino and one espresso
The goal of the task: Let the user make two cups of cappuccinos with same customisation for his/her friends. And later, make himself a cup of espresso that he saved in the previous task.
During the task they will have to perform maintenance during the process of repairing coffee. The goal is to see if the indication bar makes them aware that they need to perform maintenance soon and if the users link the lights to the icons on the display.
The redesign sees dramatic improvements in UX, Usability, and perception of a high-end machine. Navigation is much clearer and information is more supportive to understand the system.
The redesigned feature of Coffee Profile makes the coffee-making process more pleasant to most users, while the new feature maintenance forecast has a very low rate of user adoption mainly because of the failure of visual and verbal communication. The implementation of LED indication improves the finding of parts that require maintenance but its visibility (or position) can be optimized.
No testers had difficulty nagivating through the system
Coffee profile feature is easy to access and use
LED indication helped half of the testers quickly find parts for maintenance
No user noticed either the maintenance bar or its changes (due to its communication failure)
The icon of brew button doesn't make sense to 5/10 testers
Half of the users failed to see the LED indication (due to its position)
Saeco System Architecture
We mapped out the current system and visualized the interactions on a systematic level.
Below the three concepts that are tested during the analysis phase can be
found. All concepts focus on the control panel of the coffee machine. They
all explore a different way of interaction with the machine.
Concept 1 - Basic screen
This concept focuses on the primary function of coffee making. The machine has a small
touch screen with on the left and right of the
screen are two buttons for turning the machine
on and brewing coffee.
A mobile application is provided to access the
other functions of the machine, like setting up
a profile. This machine focuses primarily on
making the coffee making process quick.
Concept 2 - Full touch screen
This concept has a complete touch screen
where all functionalities are included. The
screen gives a quick overview of all the functionalities, choosing a drink, customization
and maintenance status of the machine. The
machine always gives the user an overview of
the maintenance status.
Concept 3 - Turning Knob
This concept has a more physical way of interacting with the machine. With a turning knob
to control the main functions of the machine.
There are two displays, one on the turning
knob which is showing basic information that
changes according to how you turn the knob
and what settings you are controlling. The
second display on the right to shows more
Anlaysis of 3 Concepts
- The participants liked the maintenance information at the bottom of the screen, which shows how many cups can still be made.
- The icons on the containers guided people in finding the water and ground container.
- The users were not motivated to use a mobile APP to control the coffee machine because physical interaction with the machine is needed in any cases.
- The big touchscreen made the product feel luxurious, which fits the market position of this product.
- Users are familiar with and comfortable interacting with the full-sized touch screen.
- Though there was no difficulty in finding information on the screen, it took time to read all the information on the screen before taking actions.
- People liked the physical components and the physical interaction of the machine.
- To control the machine, the knob and the buttons are needed at the same time that confused users.
Information is displayed both in the middle of the knob as well as a display on the right. it makes users confused.
There are 4 changes to the tested design, 3 to the digital interface and 1 to physical interface. They aim to communicate and hint interactions in a more explicit way as was discovered from previous tests.
Applying Machine Learning to make better maintenance predictions
Machine Learning, as a technique, has been widely adopted to improve UX. The benefits of employing Machine Learning are appealing, as it learns from users' real data and train models that can make better predictions of, for instance, users' behaviors and preferences.
It was pointed out that the rules behind the maintenance forecast we created are not trustworthy to some reviewers. However, the issue can be solved by adopting Machine Learning, but it should also be aware of the challenge of user data collection and its cost.
Increase novel smartness
It has been increasingly popular to support physical products with smart features like Philips Hue controlled by voice assistants. It for sure will be the future battleground for electronic appliances as it bridges scenario gaps, and could possibly improve UX to the next level.
Here, with Philips Saeco espresso machine, what if the machine can start preparing a cup of coffee every morning even before you get up from bed, or what if you can simply tell Saeco to make different types of coffee at the same time?