Driver assignment in on-demand logistics
Optimizing driver assignment for seamless and on-time order deliveries
DURATION
3 months (Jul 2024-Sep 2024)
MY ROLE
Research, Ideation, Interaction
design, Visual design
COLLABORATORS
Design lead, Engineers, Product team
BACKGROUND
One of our retail clients in Australia offers on-demand liquor delivery, bringing orders directly to customers' homes. They rely on gig drivers for store pickups and deliveries, using our delivery management platform, which uses an algorithm to auto-assign drivers based on proximity and speed.
Imagine a customer eagerly waiting for their order—behind the scenes, when the system fails to assign a driver, and with no safety net in place, orders get delayed and in some cases unfulfilled, leaving customers frustrated.
Breaking down the problem
We already knew what was the problem but needed answers to other W’s. For that, I noted down everything that I could gather from support tickets, stakeholder interviews, by looking at the USAT data and focused on questions like:
Who is responsible for monitoring driver assignments?
Why does auto-assignment fail?
When is auto-assignment failure most common?
Where in the journey does this issue occur?
Client side data
Understanding the "who"
In logistics, a dispatcher is a key role in an organization, responsible for coordinating and managing the transportation of goods, ensuring efficient and timely deliveries by planning routes, assigning drivers, and monitoring shipments.
When does auto-assignment fail? : Key research insights
Technical failure of routing algorithm
Drivers rejecting orders due to low payout
No driver found in a given geofence
Drivers not available due to high demand
When an order fails to get assigned or delayed, dispatchers step in the old-fashioned way—picking up the phone and calling drivers to understand the issue and re-assign the order. But this takes longer and isn’t very effective, especially during peak hours:
What happens after an order is placed?
Current experience
How might we help dispatchers take control when auto-assignment fails?
Rapid ideation
I worked closely with a mix of product managers (both client & ours) and the design lead to explore possible solutions. We ran a rapid ideation session to get all our ideas out quickly, then used dot voting to align fast and land on the direction that made the most sense to everyone in the room:
Mapping user stories
After finalizing ideas during our brainstorming session, I worked closely with product managers and the design lead to translate those ideas into actionable user stories. This helped align the team on user needs and ensured a shared direction for the product based on feasibility & impact:
As a dispatcher, I want to manually assign an order to a driver based on real-time locations, so that I can assign the closest available driver
As a dispatcher, I want to know which drivers are online and available, so I can make quicker decisions
As a dispatcher, I want to manually increase the payout for rejected orders, so that drivers are motivated to accept them
As a dispatcher, I want to prioritize urgent or time-sensitive orders, so that customer expectations are met
As a dispatcher, I need access to detailed driver information so I can make informed decisions when assigning orders
Designing the experience
As I was exploring ways to improve the fallback experience, I reflected on familiar interfaces that make complex logistics feel simple — and food delivery apps immediately came to mind.
I began sketching early concepts that placed orders, drivers, and store locations on a map — not just as a tracker, but as a decision-making tool. These early ideas eventually took shape as high-fidelity screens.
IDEA #1
Selecting orders & drivers from a list
Why it didn't work?
There’s no visibility into how far each driver is from the store in real time.
For driver status, icon indicators aren’t intuitive, and filtering requires too many steps.
Orders that need immediate attention aren't surfaced effectively.
When the order list is minimized, there’s no way to access order details directly from the map.
IDEA #2
Dragging & dropping orders with drivers
Why it didn't work?
Drag & drop feels tedious and increases the risk of misclicks during high-pressure moments
Too many icon colors make it hard to interpret, and the color-coding isn’t scalable
Order cards don’t have enough visibility into driver assignment or distance from the store
The selection snackbar overlaps with the driver status filters, reducing usability
Interaction of a hover followed by a button click on order card increases friction
IDEA #3
Selecting orders & drivers on the map
Why it worked?
A unified color coding approach — reduces visual clutter and makes orders, drivers, and stores easier to distinguish
Detailed information available within a click’s reach for dispatchers to make quick and informed decisions
Filter out and focus on urgent orders which require dispatcher’s intervention for on-time deliveries
Order and driver card - States
Assigned order - Driver rejected
Assigned order - Pending from driver
Assigned order - Driver accepted - en route
Driver - Busy with tasks
Driver - on break
Assigned order - Driver accepted - Delayed
Unassigned order
Unassigned order - Delayed
Assigned order - Driver accepted
The experience
Some future ideas
Smart Store Assignment:
We identified cases where orders could be fulfilled by multiple stores. While assigning the closest store seems ideal, driver availability can impact delivery time. We plan to revisit this in future phases. We plan to revisit a smart store-driver assignment suggestion considering all factors in the future phases.
Driver Nudges & Rewards:
To boost acceptance during high-demand periods, we will explore the idea of a gamified system that nudges drivers to accept orders and rewards them with points that unlock incentives.
Whether you're a potential co-worker, a fellow creative, or simply a curious soul, I'd love to hear from you!
Drop me a Hi and let's get the conversation started.
arushi.aa13@gmail.com
Check out more About me
Copyright 2025 - Made with love by Arushi Arora |
You can also find me on