Systems automation and model virtualisation for global network services
Single Resource Inventory Management System (SRIMS), is BT's visionary project to consolidate their network inventory – physical, logical and virtual – into a single integrated system to ensure data integrity. The focus area for this engagement was to create a data visualisation display for service network engineers to diagnose issues within the system more efficiently and to explore how 3D rendered models can assist engineers with training and issue resolution.
- Workshop Facilitation
- UX Design
- User Testing
BT and Infosys' had seen the opportunity to create this new data model which provides a unified view of the complete network covering physical, logical, service and virtual entities. Querying to find interconnected data is much faster and much more scalable, compared to the relational data model. What my challenge was to visualise how these networks could look and be interacted with by engineers in order to firstly recognise a fault and then to resolve the issue.
• Time taken to detect issues is currently lengthy and inconsistent, this needs to be improved
• Working with technical product owners who had a predetermined vision for the product
• A topology view to be created to show all service layers in one environment, these service layers can be extremely complex
In order to understand the target users we ran workshops with product owners to deep dive into their main frustrations in order to see where the opportunities were. We conducted empathy mapping exercises and explored the key user journeys we were looking to develop for the POC.
Challenging the brief
After initial workshops were run and a greater understanding of where we could add value to the user and the business. we managed to help the client look outside of a single topology view and identified how we could add value not just for the network service engineer but to take it to the next step.
"Creating an immersive environment for engineers to be instructed on how to fix the issues remotely through virtual reality integration and 3D modelling"
Initial wireframes were created and iterated to be validated by the product owner before usability testing. It was important to explore where we could integrate insights into the engineer's day through smart notifications, the ability to escalate jobs and to use machine learning capabilities to suggest solutions.
Moderated usability testing was conducted on the initial UX concepts for the selected journey of discovering and diagnosing an issue at a physical topology level. The following insights were found from the users who participated in the sessions.
User Role: Network Service Engineer
"It is much easier to see where the issue is on the home screen than reactively scrolling line by line and I can see this saving a lot of time when onboarding new engineers"
"It's a great improvement on today but it would be great to see more detail on the status of jobs by engineers, as each job is covered by several engineers but it's unclear on its current progress today"
(Efficient, useful, intuitive)
Design & Development
The insights taken from the user testing where implemented into the UX and several iterations were created for product owner approval before creating the visual design language. It was important to surface ports and connections at each service level which were shown through interaction design features and hover states.
Saving engineers and customer service time in bringing issues to the forefront, showing live status of jobs progress and any personal alerts for the user on their account to be actioned accordingly.
Creating a four tiered viewing display for the network topology, three layers can be seen in an isometric view to show their relation to one another. When the user wants to look into the detail of connections and paths they can select a single view where the network becomes fully interactive.
Creating a dynamic data model for the selected Router 7750/1 which is fully interactive, engineers can view existing connections and service ports and see where there is the capacity for new network implementation.
The next stage is to further develop these models for virtually assisted training and maintenance.