#Internet_of_things
Project "Eyes"
At Sberbank Design Center, we have developed a convenient interface for engineers and service operators of the bank.
Sberbank has a lot of real estate: bank branches, customer service offices, back offices, etc.

To monitor the condition of buildings and premises they were equipped with telemetry sensors. Sensors collect data on water consumption, carbon dioxide content, temperature, door opening, etc.

These sensors are the "eyes" of the bank. Data is collected and analyzed in a special application that engineers and operators work with.
During the pilot project, the bank's engineers and operators used the contractor's application. Our team had to rework this application under the current standards of the company, integrate it into corporate work processes. Improve the work space of engineers and speed up the response time to incidents.

We studied the documentation and the process of working with the current system. We conducted several interviews with employees. We studied the detailed business requirements of the customer. And also studied the experience of colleagues who solved similar problems.

As a result, we deeply reworked both the structure of the application and the design.
New entry point
The main change for the operator occurred on the main page. Now the first thing an employee sees when opening a service is a list of incidents and cases selected specifically for the employee.

Cases are ranked by importance and urgency. The operator can immediately respond and make some decision, or go deep into the study of details.
The incident card is another major change in the application.

If something happens at an object in the area of responsibility of the operator, the system shows him a card with basic information about the object and sensor data.

The information is submitted compactly, but so that the employee can quickly read all the necessary information.

Most of the conclusions about the cause of the incident, employees can make looking at this card.

The composition of the card:
- name and address of the object;
- the number of uncovered incidents and appeals;
- integrated information from key indicator groups;
- details of the sensors that caused the incident.
Extended information about the object
Dashboard - this is the main screen on which the operator receives information about the state of the property. We have radically reworked this section. Reduced the excess contrast, increased the information content of the first screen, Removed the information that the engineers did not use, added convenient navigation.
Each property has its own page, which contains all the information on the object.

It integrates not only information from telemetry, but also information from friendly services.

The main page of the object briefly shows all the important information:
  • data of state indicators, they allow you to quickly assess the state of the object;
  • key sensors with current data and history, which show historical data;
  • list of incidents and appeals;
  • the results of regular inspections of the facility by the service departments;
  • retrodata on sensors, etc.

Internal screens
We reworked in the new design of the system all the additional screens. For example, the object management screen has become cleaner and clearer.
Feedback of respondents
Initially, we did not plan to make a map, but the operators complained about the lack of it on research. It turned out that they are accustomed to having a visual representation of the properties in their area of responsibility.

Therefore, we reworked the cards:
  • integrated corporate mapping service;
  • added a convenient property list to make it easier for operators to switch to the "to-do list" paradigm.
Better analytics
We reworked the analytics service in the application. He lost the settings that no one used, it became easier, and began to show the data more accurately
Additionally, we:

  • integrated with various bank services: checkers, real estate database, application service, map service;
  • started developing an incident-based machine learning system;
  • Integrated real estate status display service into a corporate mobile application.

My role in the project is the lead designer. In addition to the designers, the team also worked as an analyst and researcher.

Such a team helped us make the application better: in the process, we tested all the hypotheses on real operators and engineers, we had information about related projects, and we had enough resources to assemble detailed prototypes.

Qualitative research and user feedback greatly helped us in our work. As a result, we developed an application that was highly appreciated by both users and business customers.