Solar Dispatch System

Work Order Scheduling and Analysis B2B SaaS

Project aim

Identify user problems for a new developed software that transforms from a paper-heavy to a paper-free system, and develop design solutions that help users have a better experience using the new system.


8 week internship


Donald Chou, Miranda Chou

My Role in the project

UX Research UX Design
  • Conduct user research for maintenance dispatch software system using desktop research and contextual inquiry.
  • Translated user research insights and functional requirements into design decisions and present design concepts and solutions to the team.
  • Defined visual design guidelines and contribute to a design system with unified UI components.

What is AUO and Solar Dispatch System?

AUO Solar (AU Optronics) is a leading global manufacturer of TFT-LCD committed to providing green solutions to its worldwide customers in a manner that is sustainable and friendly to the environment.

AUO provides solar solutions including solar modules, all-round solar power plant service, and integrated service platform concerning energy management. AUO's solar energy maintenance team relies on Solar Dispatch System to organize and schedule work orders across multiple solar energy sites.


How can we improve the design of the new digitalized system?

Dispatch System transforms the traditional maintenance management system into a digitalized one, as the company aims to implement ML models to improve the efficiency of the maintenance work. However, users, AUO's maintenance team, are experiencing a hard time to use the new system.


Make the workflow intuitive and reduce heavy cognitive load

User Research

I conducted user interviews and cognitive walkthrough to explore unsatisfied user needs, and did comparative analysis to understand the strengths and weaknesses of the current product.


Disparity between ideal and exact stakeholder map

This complex system is designed to serve multiple roles. Ideally, every role should interact with this system to complete their tasks.
However, the current usage situation shows that an assistant help with all the transformatioin of the digitalized work.

→ Users are not using the system as expected.

Cognitive burden during the process

When further investigating the workflow of each task, it turns out the system requires heavy cognitive load for users to complete their tasks.

→ “Recognition rather than recall.”

Part of the workflow

Design Solution

Reframe Challenge

How might we enhance the system to reduce users’ cognitive burden when creating and assigning work orders?

In order to solve the challenge, I decided to focus on developing new design solutions that encourage users to complete tasks without outsourcing the work, such as creating and assigning work orders.


Leverage digital product’s strengths to ease the burden and help users complete main tasks

Instead of asking users to remember things, the system can provide automatic information
to users to help them complete tasks without heavy cognitive load.


Assign and Schedule Work Order

  • System provides information to eliminate cognitive overload.
  • Users are able to do multi tasks on a single page.
#1 Assign and schedule order

Design concepts

A. View the schedule of executors, drag & drop
B. View the schedule of executors, drag & drop
C. View the schedule of sites with orders, click to assign
D. View the schedule of orders, click to assign

Why choose design C?

  • Fit user needs: Being able to look at unassigned orders by each site.
  • Technical limitations: Click-to-assign is more feasible and accessible than drag-and-drop.
#1 Assign and schedule order

Design outcome

Step 1: View schedule
Step 2: Assign order


Create order from AI recommendation

To make the workflow of creating orders more intuitive, I reorganized the steps and cut it down from 6 steps to 4 steps with simplified interactions.

#2 Create order from AI recommendation

Design concepts

I explored design solutions from 4 different entry points in the process.

Why choose design B?

Considering bulk actions and how the information should be automatically filled, design B is the only entry point that would not cause any data structure error and would not require users to recall information.

#2 Create order from AI recommendation

Design outcome



Designing for multiple user types

It's a great experience designing for a complex system and understanding the needs and goals for different users. Also, they are using different platforms (mobile and desktop) to work on their tasks.

Communication with Stakeholders

Working on an internal system requires gathering feedbacks from various stakeholders. Sometimes it is unpredictable when the stakeholders have availability to be interviewed. Hence, I learned to be prepared beforehand and be flexible to different situations.

To the top ↑