A fleet intelligence dashboard for a Safety-aaS solution

The process

We met Peregrine.ai founders during the ITS Congress in Hamburg, which was dedicated to innovation in the mobility sphere. After the project estimation stage, we started our cooperation.

The main goal pursued by Peregrine.ai was a fleet intelligence dashboard, something which would be convenient for fleet managers. Our client already had a working platform with an extensive risk events archive and a page with detailed risk event reviews. Peregrine.ai also provided us with the designs for a future dashboard.

First, front- and back-end developers at Lemberg Solutions created a dynamic dashboard with a large number of graphics displaying vehicle statistics. The graphics included average risk score, risk components, risk factors distribution, trends overviews, as well as the top three vehicles involved in risk events and those that reduced such risks on the road. In addition, our team created filters for reviewing the statistics of a particular vehicle or a group of vehicles.

Second, our developers worked on a detailed page consisting of a chosen risk event video, a map showing the exact location of an event, and a graphic to which we added a runner that moved simultaneously with the vehicle on the video and showed detailed information about the event. On top of that, we added a feedback loop feature that allowed a driver to evaluate the level of hazard on a specific part of the road.

We also added specific functionalities to the product, such as video deletion and archiving features.

For front-end development, we applied React and GraphQL. The LS team created an in-between environment for transferring front-end requests to the back end. Our engineers used Golang to extract data from the client’s database as per request from the front end, and we also created a logic for the data extraction and transfer.

Since the product is multilingual (German and English), a number of titles on the graphics weren't displaying properly, and our team undertook fixing this issue.

The LS QA engineer tested the product during the whole development process to make sure all statistics and requests performed correctly.