Paria Adib's profile

Smart Locker Web App - Case Study



Overview

Ganje locker is a smart locker that receives parcels instead of users.
To use Ganje, users have to connect to lockers through a web application that connects locker and user and helps them to pick up their parcel from lockers.
Ganje lockers are secure storage and distribution systems with integrated computers and sensor networks that provide users a self pick up delivery.
Through a web application, users are able to connect to lockers and pick up/ drop off their order shipments.

Problem statement

According to field observation and sensor logs, we noticed that sometimes, users didn’t close the locker doors after picking up their parcels. This matter caused:
Next users/ Couriers couldn’t recognize their reserved locker.
It might have a bad impression on passengers who are not direct Ganje users but are in the neighborhood.
In big stations (e.g locker racks above 20 doors) or noisy outdoors it was too difficult for the users to find out which locker door was opened for them.
Based on the CRM calls analysis, we find out that when a locker door pops open, users were not able to find the door, so they end up calling the call center and they would guide them to their locker based on their reservation info. So the question was why users couldn’t find their door although it opened after the picking up confirmation on the web app.

Getting to know the problem

Field study
To solve the problem, first we needed to know why and how often this case was happening.
To answer the why part, we investigated the station area and also, observed users behavior while picking up their parcels.
The observation conclusions were:
- Noisy environments lead to not hearing the door opening.
- The door was not opened completely.
- The last, and most common reason was the doors left open by the previous user. So the user reached out to the door that was already open and ended up with an empty locker.
Quantitative research
To measure how difficult it is to find the opened door, we run a survey on our web application and ask our users did they find the open door easily?
We noticed about 12% of users couldn’t find their parcel door (According to the statistical society who participated in the survey, we can be sure about 95% of the results).
The CRM team also confirmed that every day, they receive around 30 calls, users complaining about this issue. This is about 10 % of our orders per day.
Asking users if they find the opened door easily. About 87% of users said they found the door easily, but almost 13% said they have troubles finding the opened door.
To be certain about the frequency of the problem, we investigated sensor logs and found out that about 13% of the lockers door didn't close after popping up.

Chopping the problem into smaller pieces

So with all this shocking data, I decided to chop the problem into smaller pieces and investigate them separately.
Problem #1: users couldn’t find their reserved locker
At the beginning, the solution was easy; show them their reserved locker! But we have different ways to do it and we had to decide which one is the best for us:
1- Add LED or labels on the lockers in order to show users the open door
2- Show the station on our web application and emphasize on the reserved locker.
Knowing the fact that we had 25 stations ( about 100 racks) in Tehran, we were sure that any physical changes could cause a lot of operational work. Also, our locker faces work as billboards and businesses rent the face for marketing goals. So it would have been better that we had a solution that didn’t make too many changes on face.
So, we decided to go with the solution #2.
Showing users the open door with a locker simulation, plus addressing the opened door with text
Within two weeks after launching the feature, the number of CRM calls dropped about 14 % .
Also we ran the survey again and as the results showed that the problem was decreased to 2%.
Problem #2: Users didn't close the door after they picked up their parcel.
Through more investigation, we noticed that the couriers sometimes come across the same problem too. When a previous user doesn't close the locker’s door properly, the next courier won’t be able to find the locker that they must drop off parcels there. An open locker can cause miss understanding to them and they might put the parcels in the wrong lockers. In that case, a lot of operational work has to be done in order to resort parcels in their assigned lockers.
Therefor it was vital for us to make sure that the locker’s door is closed when any type of user comes to lockers.
We had some physical changes in our mind, like a speaker for alarming or a leverage that pulled the lockers door to close, but it cost a lot and needed a lot of operational work. So, we decided to use other touch points through the web- application.
Alerting users about closing the door
We wanted users to be sure that they had closed the door after picking up, so we tried alerting them on the opening door page, which led to changing the design of that page. As a plan B,( in case this alert wouldn’t have the effect) we would have used a sound alarm on the web application, but the rate of doors being open dropped from 13% to 5%.
Furthermore we changed our courier web application in a way that there would be less mistakes in loading lockers. At the first of their journey, we show them open doors and ask them to close it. Also for every parcel, they would see which locker is going to be open and hope that it would decrease confusion.
Showing open doors to couriers and ask them to close them before they start to work with Ganje Lockers.

Outcome

The most interesting part of the job to me was watching the impact of my design on users and the company.
With the changes I made, the rate of our CRM calls that relates to this issue, dropped significantly. In addition, solving the open doors problem prevents us from having a bad impact on passengers that see our lockers with open doors and think it’s not safe enough to use it.
Smart Locker Web App - Case Study
Published:

Smart Locker Web App - Case Study

Published: