Fly Buys Drone Delivery Service was an App designed for a uni subject in which we were required to design for Autonomous systems and explore how we could create meaningful human-machine interactions. This App in particular facilitated an Autonomous drone delivery service which tackles the frustrating delivery experiences of current day apartment residents. The App integrates into existing e-commerce sites in a similar fashion to 3rd party payment providers such as PayPal or Afterpay. Once the user has set up their delivery and paid for their product they are then onboarded onto the App for tracking and receiving their delivery.
For this project I worked across the board from the initial UX Research, the Ideation stages, Information Architecture, Wireframing, User testing, and High Fidelity UI design stages.
We began by conducting extensive user research with our goal to find out how people living in apartments receive deliveries and how happy they are with their current services. We used a combination of qualitative and quantitative methods to gather our data; namely Interviews, Surveys, and Direct Experience Storyboarding.
After analysing and synthesising our data we were able to understand the key user needs that apartment residents had:
"I need delivery options to fit around my work and life commitments"
"I need to not have to leave my apartment when my delivery arrives"
"I need to know that my delivery/parcel is safe when delivered"
"I need clear, accurate real-time tracking and communication about my delivery"
"I need to not have to drive to pick up my delivery"
From this we were able to develop our problem statement:
"How Might We create an apartment delivery experience that is flexible, convenient and secure for apartment residents?"
After formulating our research findings we turned our attention to ideating solutions that would adequately meet our problem statement and solve all our key user needs. We developed 3 different concepts and storyboarded these whilst providing pros, cons and considerations for each of them. From this we were able to evaluate our concepts and choose the one that best fit our problem statement and user needs.
Chosen Concept - Drone Delivery to Balconies
Our chosen concept was a drone delivery service that allows customers to have drones land on their apartment balconies to deliver goods. These deliveries would be accessible after business hours and would arrive within short windows of time to provide a service that is convenient and reliable.
After choosing our concept we began exploring the information that users would need at each stage of the journey. We realised that the delivery experience was linear and could fall into 4 different stages. We used card sorting to map out the information architecture and how it would fit with these 4 stages.
Low Fidelity Prototyping.
Our Low Fidelity prototypes began as very rough, quick sketches and progressed to nicer and more detailed paper prototypes that we used for our first round of user testing.
1. Low Fidelity Sketches
We decided to start thinking through the more complex and unique functions that our App would undergo. A lot of these functions weren't unique in and of themselves but were made more complex due to the autonomous nature of our drone delivery technology that we were designing for. For example the traditional delivery experience only requires users share a residential or postal address. Autonomous drones offered exciting new possibilities such as getting a delivery to your balcony or somewhere else in your apartment complex. So this required new design patterns.
2. Paper Prototypes
We developed the paper prototypes in order to test the two different user flows we had come up with. The paper prototypes were drawn on paper and then entered into the Marvel app via photo. This allowed us to create ‘clickable links’ so that users could click the main CTA buttons and understand the flow of the screens.
User Testing 1.
The goal for our first round of user testing was to explore which of our two user flows provided a more intuitive user experience. Our user flows differed in the sense that they placed the requirement for the user to enter the exact delivery location at different stages. Flow 1 placed this event before the commencement of payment i.e. while the user was still on an ecommerce platform and ordering their product. Flow 2 on the other hand envisioned that this would take place later once the parcel had been ordered and as the user was setting up the drone delivery app.
Method: Usability Testing with Think Aloud & Observation
We set the users a number of tasks to perform with the paper prototypes in order to simulate the functionality of the app. These had been put on the Marvel app in order to provide the realism of using a smart phone. One set of our users tested Flow 1 first and then performed the same set of tasks with Flow 2. The second lot of users started with Flow 2 and then went on to use Flow 1. As they were performing these tasks we encouraged them to use the Think Aloud protocol and verbalise their thoughts, feelings and frustrations. On top of this we observed how well they were completing these tasks and whether they had any difficulties or if their actions were contradicting their verbalised thoughts. Once they had completed the tasks we asked a few semi-structured interview questions aimed at getting the user to reflect on their experience.
Throughout the first round of testing we took notes and also an audio recording of the session. We were then able to go away and analyse the data to get our key findings:
After receiving feedback from our first round of user testing we iterated on our paper prototypes to make improvements based upon the results we discovered. At this stage we chose to increase the fidelity of our prototype and created digital wireframes using Sketch. These were then added to Invision so we could create clickable hotspot links and connect all the screens. This was done so that we would have a more realistic prototype that we could completely click through for our next stage of user testing.
User Testing 2.
At this stage we wanted to focus more specifically on the usability and intuitive nature of each screen and the UI elements and patterns that the user encountered. For example in the first test we were more concerned with the big picture, the overall user flow and when we should get the user to select the delivery location. Here we wanted to focus on their experience of actually selecting the delivery location for example and make sure this experience was coherent and simple.
Method: Usability Testing with Semi-structured Interview
We had the user complete a lot of smaller tasks i.e. ‘choose the delivery time, choose the delivery location’ and then asked them questions straight after they had completed each task. In total we broke the App into 20 different stages and asked a set of questions at each of these stages, so it was quite an in-depth process. These questions were particularly aimed at gauging their thoughts, what they liked, disliked and any concerns they had about each screen and the functions that they performed. We took notes at every point that we asked questions. After we had done this for every stage of the App we asked some brief reflection questions aimed at getting them to reflect on their entire experience and the overall flow of the App.
During the second round of user testing we took notes and also an audio recording of the session. We were then able to go away and analyse the data to get our key findings:
High Fidelity Prototype.
After receiving feedback from our second round of user testing we iterated on our wireframes to fix the remaining issues. From here we were able to start developing the visual style and branding in order to develop our high fidelity prototypes.