Sector
Data Analytics in Mining
Anglo American
Project Time
5 months
My Role
Research,
Ideation and concepting,
Stakeholder engagement,
Synthesis,
Wire frames and Journey mapping,
Ui Design
Users, Groups, Roles and Applications
In this project I was tasked to revise the experience of an authentication system that initially was seemingly complex. Through research and synthesis, I uncovered that although technically complex in the backend, what the end users were trying to get to as an outcome could be achieved in a relatively simple and manageable way.
I worked in a team of 2 to create an efficient system that improved and optimised the workflows of technical managers, saving them huge amounts of time and effort.
The Challenge
The first step in this project was in defining what the actual problem was. I facilitated the discussions and workshops with key stakeholders, most of whom were busy site managers and lead tech support engineers who were time poor and looking for a way to simplify and streamline a process that was taking them far too long and yet should have been a relatively straight forward task.
We held many workshops to understand the complex nature of the work of our users and the problems they were trying to solve on a daily basis, particularly in relation to the use case they had with this permissions system.
What we found was that a big part of what they were responsible for on a daily basis was the administration of authentication to many, sometimes hundreds of users across different mine sites and the configuration for those users, their roles and what they had permission to access based on their roles.
Their current workflow for administering access and controls to users meant that there was a lot of repeated process and action and they couldn't delegate the work to other people. They needed a system that could control and automate this.
Problem Statement
"As a Mine Site manager, I have spent 3 weeks setting up user permissions for Mine Site 1 and now I need to do it all over again for Mine Site 2. I want to use the same roles at Mine Site 2 as I did at Mine Site 1, except I need to change the parameters based on the new locations"
Research Responding to Partial or Incorrect Information
We knew that the existing system was getting too complicated.
The more the tech team tried to facilitate bringing in different functionality, the more it became unclear who would be using the system and what they would really get out of it.
So we paused and looked at our problem statement. Distilling this, we have:
1. Site managers who are time poor
2. They manage permissions of large amounts of users
3. They have specific parameters they need to administer but they also are replicating those parameters across sites.
We understood the system had 4 primary components to it:
1. Groups were there to act as ‘buckets’ of many users
2. A user was a person who needed to take an action anywhere in the organisation
3. A role was a set of empty functionalities that could be assigned values as it was given to users. For example: “This user can press this big red button” (functionality) at the mine site Kumba (value)
4. Application referred to the specific product where the action would be carried out
We were essentially building an automation tool for our colleagues to save them time and money for the business. This was a great position to be in because it meant we could bug them with questions and try to get an understanding of what they were doing every day and what the number 1 priority within this system would be for them on a long term scale.
What We Heard
“If I can just take these same sets of permissions and roles from this existing group and bring it over to a bunch of these other users, I think that would save me so much time”
SynthesisAfter relaying with our colleagues, our end users several times, we got to a place where we understood the nature of the issue very well. I also worked with developers to understand the backend architecture of the system and how it could be plugged into all applications across the organisation and how it had power in automating the process of administering users.
In this instance we got lucky, and because we were working with domain and field experts, we didn’t need to spend too long observing and testing because our users in this case were very well aware of the problem and what would need to happen. They just didn’t know HOW to make it happen or how to make it accessible and usable. Which is where we came in.
Normally at this stage in the UX process, we would spend some time in creating a persona to capture the voice of the user and have that as a guiding light house to refer back to as we went through the design process.
Because we worked so closely with these users and the problem was a relatively well articulated one, we felt confident that we could take the existing work the tech team had done and re package it to make it usable.
PrioritisingWe knew there were multiple pain points but we needed to segment this so that we could deliver incrementally and make the most critical tasks our priority before moving on to the ‘nice to haves’.
After working and validating with the tech team we identified 3 killer features that needed to be prioritised in order to get users using the system and giving them back their precious time.
UI flows and wire framingWe set about creating UI flows that showed the different actions that users needed to take at each step of the journey and on each page of the system.
After spending a day or two in understanding the technical, backend processes, we were able to create a UI flow that gave us the structure to start building out screens.
This ensured that at every screen and every step of the users journey, they had the right functions and access they needed to automate the process of assigning roles and abilities to as many users as they wanted as many times as they wanted, with as many different values as they needed.
High FidelityAfter ensuring we had accounted for all the right steps, we set about bringing the flows to life through high fidelity designs.
At this time I functioned primarily as Anglo American’s design system manager which at that stage was quite robust.
This was particularly beneficial because it meant very little re-work in building components and the overall UI of the system. I didn’t have to create anything from scratch and it meant we could jump from wire frames to high fidelity design almost immediately.
Testing and sense checkingWe conducted somewhat informal usability testing with our users as we were already having daily stand ups with them and they were using the prototypes as we built them and submitting feedback along the way.
As we were testing with power users of this system and similar systems, it was particularly useful and insightful to be given direct feedback around the parts that worked and what didn’t work and how we would be able to achieve 100% usability and satisfaction in a small time frame and with little rework.
ResultsWe haven’t implemented the analytics tools that measure time on task, but from a qualitative perspective we sent out surveys to 8 users that asked for responses in user satisfaction, confidence and ease of use.
The scores told us that we had succeeded in what we had set out to achieve. From the results of the survey, the usability of the system scored 99% for success on tasks, 92% for confidence and 97% for ease of use.
Additional to this, the team we built the system for requested we give access to other mine sites and areas of the business as the impact of using it was so positive.
Teams across Anglo American started adopting the system and because our next focus on the business was in building a global notifications dashboard for alerts and reporting, I handed over the reigns to another designer to scale the system as more feedback came in.
I was particularly proud of this project because we solved a problem that many other developers of enterprise products had not yet done. Working in enterprise level design, we work with many 3rd party tools and this was an instance where no one had cracked it and we had the opportunity to roll up our sleeves and figure it out for ourselves.
