MEAN Stack Developer

About Urban Sensing Ltd

They are an innovative and highly technological organisation serving the ‘Internet of Things’ sector within the global Smart City market. Urban Sensing Limited is an IoT solutions manufacturer providing critical elements required for a complete Smart City Solution. IoT Data is provided through our UK manufactured products capable of being fully integrated into various data analytical smart city platforms.

They provide indoor/outdoor footfall, automatic passenger/vehicle counting and data acquisition sensors for smart cities

My Role

Armed with NodeJS, ExpressJS, MongoDB Atlas, Amazon Webservices (AWS), D3.JS and whole 'lotta learning, I was tasked with designing and building a full web application. Including back-end configuration, security, user accounts and authentication. Pages needed to be dynamic, meaning client users would only have access to their data, with graphical formats such as bar, line and pie that dynamically update to real-time data and filtering.

Below are several examples of such work.


Due to the sensitivity of the data used on the Urban Sensing application, and legality issues concerning The Data Protection act (1998), all the data used on this page has been generated by a script to emulate the data model of the real page. You can learn more about how I generated it here.

Currently Displaying





< 33% of range

range > 33% and < 66%

> 66% of range

Toggle Passengers In/Out

Journey Information

Weather Information

Time Recorded


Weather Type

Wind Speed

Wind Direction

Data Visualisation

Retail Sensing Camera Comparison

The first page I worked on under Retail Sensing (this was after I finished the Urban Sensing project) was the comparison of data from multiple cameras in a building. I worked on this with fellow colleague Peter Jedra. We had to create a script that read data from '.wl' files and committed it to the database.

Select Business

Select Month

Update Graph

Customers In/Out

Clear Graphs

The line chat below shows the number of customer that walked into a number of different rooms. The back-end algorithm has been reduced a lot, but in practice was much more complex due to the file types and data structures we had to parse and build.