Kompano, the Autonomous Greenhouse Robot

Redesigned app for operators to control the autonomous robot in a greenhouse by providing an overview, simplifying interactions, and making them feel in control

The challenge
Kompano, an autonomous robot for cutting tomato leaves in greenhouses, was developed by Priva to address the lack of skilled workers in the horticulture sector. The Kompanos are operated through the Kompano Mobile App, which is used to start and stop the Kompano and notify when errors have occurred. However, app usability needs to be improved to enhance time efficiency and prevent financial losses.
Value delivered
Our team evaluated the usability of the current Kompano Mobile App and identified the main problem that the operator lacks an overview of the tasks to be done and how the functions of the app work, which leads to a feeling of lacking control.
 
We redesigned the app to provide operators with a clearer overview of the location and status of all active robots. Testing the redesign against the old design resulted in a significant 54.4-point improvement in the System Usability Scale (SUS) score. Users also completed tasks, on average, 20.6% faster with the redesigned app compared to the original.
My contributions
📊 Analyze data and draw insights
✍️ Visualize problems with storyboards
💡 Ideate interaction concepts
📱 Prototype a new design with Figma
➡️ Structure the project’s logics for reports, questioning “why.”

Category

Academic / Team

Type

Product / UX / UI

Scale

Robot : 1910mm(L) 880mm(W) 1800mm(H)

Period

Feb. 2022 – Jun. 2022
The course from UXAD (User eXperience Analysis and Design)

Role

Design Research, Concept, Prototyping, Evaluation

Client

PRIVA

Worked with

Kristin, Lea, Natanya, Celine

Supervised by

Arnold Vermeeren
Sergej Schetselaar

What is Kompano?
Source: Priva

The robot designed for deleafing tomato plants in greenhouses

It cuts back the older leaves from the tomato plant to provide more sunlight to the tomatoes and facilitate harvesting. The Kompanos are controlled and monitored by the Kompano Mobile App. An operator uses the app to scan for nearby robots and connect with a robot via Bluetooth to operate it.

More information about the robot is H E R E .

priva_logo2

Priva is a well-established firm leading the market of high-tech products for the horticultural industry. The Kompano is the first solution for a fully automated deleafing system, representing a significant achievement and advantage for the company. However, it also represents a challenge in controlling the app connected to the Kompanos because of its experimental technology, which is easily subjected to technical issues and unexpected malfunctions.

Project Overview

First-hand Experience

Understanding the context through Field Research 

Storyboard of the current workflow

Usage Inspection & User Tests

Cognitive Walkthrough

Critical contexts:

  • When operators starts a Kompano
  • When operators notice errors and try to fix them

Finding 1.

Lack of Overview

Differentiating the names of Kompano was challenging because the code names looked similar. Moreover, to fix the errors, the users needed to go close to each Kompano in person.

Finding 2.

Unintuitive app flow

The inconsistent bottom tab confused users. In addition, the terminology was confusing for users to understand what the features exactly do. This led to inconsistant affordances to users.

Finding 3.

Confusion when handling errors

Error descriptions showed insufficient instructions to solve them. In addition, the feedback lacked as some users tried to touch some icons because some were not clickable. Even when the problem was solved, the feedback was unclear.

Finding 3.

Confusion when handling errors

Error descriptions showed insufficient instructions to solve them. In addition, the feedback lacked as some users tried to touch some icons because some were not clickable. Even when the problem was solved, the feedback was not clear.

Scoping the Problem

Problem Statement

The operator lacks overview to operate the Kompanos efficiently.

Locating the robots

Understanding robots' status

Fixing errors

Definition of ‘overview’
  • Location and status of the Kompanos
  • All actions that the operator needs to perform

Design Goal

Provide the operators with
1) an overview, 2) simplified operator-Kompano interactions, enhancing operational efficiency, and 3) a sense of feel in control.

Conceptualization

We used Crazy 8 (from one of Google’s design methods) to create a new concept of redesign. Moodboards were used to visualize “feeling in control”. With the chosen moodboards, Crazy 8 was used to come up with as many ideas as possible in limited time.

Simplifying the hierarchy of the menus on the top and bottom

AS-IS

TO-BE

Maps for visual communication

The map provides the operator with an overview of the location and the statuses of the Kompanos. Since every greenhouse has a different layout, the map is highly simplified, so it displays only the primary path and a compass to indicate the direction. The map has a blue dot to mark the operator’s position.

 

List of all the Kompanos before any filters are turned on
Out of battery filter is on, showing only the Kompanos out of battery
Kompanos with different statuses
Additional status information when toggeled

List for a quick overview of the robot’s status

To quickly find the Kompanos with the same status, the app has a filter function to group all Kompanos that are running, out of battery, have a consumed knife, etc. The order of the filters is sorted after occurrence, providing a hierarchy of the most common statuses. In the list, the Kompanos are organized by an algorithm based on priority of intervention. The algorithm will take into account the position of the Kompanos compared to the operator, and the urgency of the error.

Visualization of the program

The program of the Kompano is based on moving forward and backward, cutting on the left and the right side. These parameters vary according to which side of the Kompano the operator is standing on the path. These settings are visualized here to make it more intuitive for the operator to understand the current program.
The visualization relies on the Kompano being physically slightly different on each side. The Kompano will have a yellow sticker on one side and a blue on the other.

Guidance in handling errors

Error handling is designed as a step-by-step guide to increase the operator’s feeling of control. The step-by-step guide is beneficial for new operators and operators who are less fluent in English, providing them with a visual explanation of what they need to do.

Design Validation

54.4 point improved in usability (The SUS Score)

Having an overview

Time efficiency increase​

Feeling of being “in control”

Time efficiency increase​

  • The number of times that participants did not click the correct button decreased by an average of 7.1 times.
  • Clicking times on a non-interactive element is decreased to an average of 3.1 clicks.
  • Participants got lost on average 0.1 times more in the redesign.
The amount of times participants click wrongly, get lost in the app or they misclick

Takeaways

Improve unclear visualizations

Some participants did not understand how the visualization of the program works. They also did not realize that the arrows were clickable. The test shows that the user will need more time to get used to that type of visualization and learn how it can be managed. 

In addition, the map has room for further development. The map could be standardized in different green houses. Due to greenhouses having different layouts, the map is limited to showing the main path, the position of the Kompanos, and the operator.

Inclusivity

We have focused on conveying visual information to make the app easier to navigate for people with limited language knowledge. Since the user group consists of multiple nationalities, inclusivity is an important matter.

Further testing

I learned that pre-testing helped our team to get the most out of our user testing before conducting it with participants. We could enhance our testing to the next level by conducting pre-pilot and pilot testing of the project. A benchmark test is also practical to ensure that our data is as comparable as possible. This has given more value to our results.