Barcelona, June 2018
At Worldsensing, the Engineering team uses Python programming language to visualize, analyze and predict parking occupation, amongst other data sources. To share our experiences with real-world IoT data, Worldsensing hosted an event “Data science for smarter cities” for the PyLadies BCN meetup group.
We are entering the era of megacities. Today, half of the world population lives in urban areas, and this proportion is expected to increase to two thirds by 2050. This situation brings a wealth of new opportunities, but also a number challenges. It is estimated that about 30% of inner-city congestion is caused by drivers looking for a parking space. The key is to use smart technology to obtain information on parking behavior and availability in real time to optimize urban operations and prepare cities for the future.
At the workshop hosted at the Worldsensing headquarter in Barcelona the participants analyzed smart city parking data to get insights into how parking behavior evolves over time. This workshop was the first PyLadies BCN event on smart city technology. PyLadies BCN is a Barcelona chapter of an international mentorship group aiming to help women become active participants and leaders in the Python open-source community.
Using real smart parking data obtained from Worldsensing’s Fastprk2 innovation project which has led to the development of the recently launched Fastprk Evolution parking suite, the meetup participants visualized and analyzed parking occupation data with the aim to improve parking guidance for drivers and minimize time spent looking for a parking spot.
While the Worldsensing engineering team uses Python from start to finish to develop the company’s own Smart City and Industrial Internet of Things (IIoT) products and solutions, the PyLadies workshop focused specifically on how Python is used with Big Data. Worldsensing engineers use Python to analyze, craft and deploy models onto production, and train those models to predict traffic flow deviation and parking occupancy by using Machine Learning and Deep Learning techniques.
The PyLadies meetup was hosted by Jamie Arjona, an in-house Worldsensing Software Engineer.
“I’ve always been interested in the information that data can bring to us, and how it can be extracted and used. What attracted me most to the field is the possibility to model the information and create mathematical functions that emulate the patterns hidden in the data.”, Jamie reveals.
Jamie is currently working on his PhD which is made possible through a collaboration between Universitat Politècnica de Catalunya (UPC) and Worldsensing. As part of the testing phase of the Fastprk2 innovation project in 5 EU countries, parking sensors were installed in the UPC parking lot in Barcelona to conduct a field trial of the technology. The objective of Jamie’s thesis is to create mid-term forecast of parking availability using historical sensor data in order to modelize the parking patterns that occur over time. In his role at Worldsensing, the Software Engineer uses Python for the whole pipeline: from extracting data from the data warehouse, to transforming, visualizing and modelling it.
Today, Worldsensing’s smart parking technology, Fastprk, predicts parking occupancy based on deep learning.
“What connects deep learning and smart parking is that with enough data, any continuous function can be modeled. Parking availability is governed by a “hidden” function that can be learned with enough data and the correct observations.”, Jamie explains.
For more information about how Deep Learning is used to predict parking occupancy, the Worldsensing innovation team published an article on Deep Learning and Smart Parking.
Worldsensing is a widely recognized global IoT pioneer. Founded in 2008, the Barcelona-based technology provider delivers Operational Intelligence to traditional industries and cities. With over 90 employees and offices in Barcelona, London and Los Angeles, Worldsensing is globally active and has customers in over 50 countries across 6 continents.