Taiwan Taxi Company
The Drivers' Lift Feature
Taiwan Taxi Company is the largest taxi company in Taiwan. The company serves about 350,000 customers with more than 10,000 drivers running taxis.
This is a 0 to 1 project which focused on improving drivers' experience on the mobile app and helping them gain more income based on the AI model.
3 months, 2022
UX Designer, Miracle.ai
Mobile App, Transportation Network
The drivers can easily tap the Home button and the AI model can help them estimate and find passengers who have a similar route.
The drivers are also able to modify the destination in addition to their home.
They are able to edit the current location and destination as well.
Voice control also provides another safety option when they are driving.
The Distance button allows
taxi drivers to define how far
they want to drive for accepting
new ride orders.
Dragging the milage slider can achieve the goal.
Taiwan Taxi company provides excellent transportation service to customers and has won Taipei Taxi rewards for consecutive 7 years.
The company wants to continue to maintain its good image of the brand and also support taxi drivers economically.
I started the design for the key users: the taxi drivers.
Taxi Drivers
"It's time to go home!"
"Eek, it will be nice if I can still take a ride on the way home. I can make more money then.
But... how can I know if there's anyone who needs a ride? "
Based on research, there are several key pain points that the drivers have.
The taxi drivers want to gain extra income by giving a ride to people on the way home.
Current Problem
A majority of people do not tell directly where they're going before they get on the taxi.
There are only 8 seconds to decide to accept a request which makes drivers feel rushed and frustrated.
So...
How might we help taxi drivers gain more income?
How might we help taxi drivers take orders confidently?
Gain extra income by giving a ride to people on the way home?
How might we help taxi drivers gain more income?
I designed a new Lift feature that taxi drivers can take orders on the way home or the other destinations.
A feature called "Lift" was incorporated into the original app so that a taxi driver can tap on the button and start taking orders when he or she is on the way home.
On the address card, the driver can use the preset home address by touching the "home" button.
There is also some other frequently used destination that drivers can choose with one single tap.
The system, incorporated with the AI model, can predict people's destinations and show those who
have similar routes.
With the user testing, I got feedback from the taxi company and drivers.
So I did iterations and add a new feature.
"I feel safe using the home shortcut.
But I do not know where I am now and do not want to see people throughout the whole map.
Is there any way that I can filter it?"
A "Distance" filter was integrated so that drivers can define the range of people they take the order.
Then the home address will be automatically filled out.
.png)
A selected group of people who need a ride is seen on the map.
people do not tell directly where they're going before they get on the taxi?
How might we help taxi drivers take orders confidently?
We did the explorations and incorporate AI model to help predict users' drop-off location based on previous location history and data.
After entering the address, AI Model can search for and show people who have a similar route as the driver's.
There is a purple " people" icon
shown on the map.
various degrees of color saturation indicate similarities.
The denser a purple person is, the more likely a person will be on a similar route.
A " The Smililarist" Label will be shown on the map to help drivers identify thus reducing cognitive load when on the road.
"How can I take orders?"
"What if I am not satisfied with
current recommendation?"
Based on the AI model, the system can also give recommendations to passengers.
It shows the predicted drop-off location, AI matching score as well as estimated taxi fares.
Drivers have the freedom to check on another recommendation by taping the "Next" button.
Going back to the previous address card is also an option by taping on the "Location" button.
The Taiwan Taxi Company adopted this feature and is integrating into its current mobile application.
Metrics
how many drivers use this Lift feature?
how satisfied are the drivers with this new feature?
Future Development
Continue to Collaborate on prioritizing drivers' using experience of its mobile app
Using AI Models to optimize
the "Lift" order recommendation
based on "Distance"
Using AI Models to optimize
the "Lift" order recommendation
based on "Time"
Using AI Models to optimize
the "Lift" order recommendation
based on "Fare"