Brighter Spaces: Intelligent Lighting

Different systems in buildings with their sensing capabilities create a vast amount of data. Yet, most of the buildings are not that good at creating holistically good conditions for the users of the space. When analysed, applied, refined, and visualised smartly, sensor data can help understand a buildings' usage much better. This creates several opportunities to build data-driven services when a building is being taken into use and during its lifetime for technology-specific use cases, e.g. lighting or HVAC, but also allows to build a holistically great user experience for the truly intelligent building of the future.

The challenge

In this challenge, we ask you to utilise, apply and visualise the sensor data to understand the use of the building and to improve applying the sensor network optimally in the building in the first place. You can select one or all three of the parts of the challenge to work on.

Where are all the devices?

We typically spend a lot of effort on identifying and positioning a large number of sensors on a floorplan. Can we simplify the process, and use the data from a subset of the sensors to automatically place the rest on a floorplan?

Where is everyone going?

Motion sensors are useful in determining the amount of activity in a space. We also know what kind of space it is, for example, a hallway or a corridor. Can we identify what are the most used pathways in a building, given the hundreds of thousands of motion events the sensors are generating?

Mystery Garage

There are moving vehicles and pedestrians in a garage. We have motion and audio sensors. Can we identify what is happening in the garage? Are there vehicles coming in or going out? Are pedestrians walking towards the elevators, or getting in their cars?

Insights

What We'll Bring

We will provide layouts of the different floorplans containing device locations and time-series of motion and audio events.

Datasets:

BUILDINGS The participants are given datasets from indoor premises. The data will be completely anonymised. The floorplans will contain devices and their locations. The dataset will be a time series of motion events from all the sensors on the floorplan.

GARAGE The participants are given a training dataset from an indoor garage, collected using motion and audio sensors. 

You are encouraged to use other data sources and feel free to use data from/combine your ideas with other Junction challenges if you see a fit.

The Prizes

Prizes will be awarded for the top 3 submissions —we will select the three best solutions out of all the participating teams regardless of the challenge. Each of the winning teams will receive a €1000 Amazon gift card. From the three winners, we will select the overall winner of our Junction challenge.


The prizes are handed to the recipients virtually as Amazon.com eGift Cards.

Judging criteria

We would love to see quality Machine Learning and/or analytical approaches to the challenge, but we encourage participants to come up with out-of-the-box solutions.

About the Company

At Helvar we create intelligent and energy-saving lighting solutions. Intelligent lighting is essential to the wellbeing of employees, customers, visitors and patients, and it helps to achieve sustainability targets. Helvar lighting control solutions help to create smart environments built for the future, using information and working seamlessly with other systems. With a wide range of lighting controls and luminaire components, we bring open and flexible solutions for lighting and smart buildings designed with people at the heart. Worldwide, we serve our customers locally by our sales offices and our global network of partners in over 50 countries. To learn more, visit www.helvar.com.