SIMPaCT Smart Cities

Smart Irrigation Management for Parks and Cool Towns (SIMPaCT) is a research project designed and led by Western Sydney University with Eratos, New South Wales Government, Total Water (water supply manager), and other parties concerned with machine learning, hydration models and sensor provision.

The project has been set up to illustrate how soil moisture sensors, weather stations and various federated data sets can be used together to provide intelligent irrigation to maintain healthy and cool recreational areas and parks.

These types of solutions are becoming increasingly important as global temperature increases impact our green spaces and water usage dynamics. As global citizens, it is our collective responsibility to utilise our scarce resources responsibly and efficiently.

Challenge

The absorption, storage and radiation of heat by grey infrastructure leads to urban warming. Called Urban Heat Island (UHI) effect, this creates warmer urban environments during extreme heat events. SIMPaCT can help reduce the UHI by generally reducing air and surface temperatures in and around the park during warm and hot summer days.

In order to understand the current state of the park and then build a solution to apply the appropriate mitigation measures, various, disparate data sets are required to be interrogated together. This, along with assembling bespoke workflows and building fit-for-purpose machine learning models, the project required all of these components to be accessible in one platform in order to build the desired solution.

With the Eratos platform, a hub-and-spoke approach is possible with Eratos representing the central point to allow data and models to be seamlessly accessed whereas without a central point, this is challenging to manage.

Solution

Eratos is a Data Infrastructure Platform solving the way organisations and individuals make informed decisions about the world we live in. Eratos makes finding and accessing disparate data at scale effortless, with interoperability and reusability built into its DNA.

Eratos does this through three products:

a data infrastructure platform to make any dataset available in situ (PaaS);
a developer environment for Creators to build models, workflows and applications (SaaS); and,
a marketplace to publish and share scalable solutions built on top of the platform.

On Eratos, you can access, build, collaborate and share data, models and solutions. Particular focus to date has been on research, education, federal and state government, agriculture, financial services and smart cities.

More than 250 sensors were deployed across the parkland to record ambient environmental conditions. These included soil probes, tree sensors and weather stations. The Eratos platform, incorporating Senaps, provides the hub for all data feeds from these sensors.

Models and Machine Learning are incorporated into the solution to provide appropriate modification to the irrigation scheduling to maximise the cooling effect of the irrigation system. These models and machine learning algorithms are built and hosted on the Eratos platform.

The images below are map screenshots from the SIMPaCT live dashboard called Park Now. The maps interpolate sensor data from the last hour in the form of a temperature map (first image) and soil moisture (second image). The sensors also return data of their location and the contour of their landscape within 2 metres.

Leveraging technology partners

The SIMPaCT solution leverages the Eratos Senaps platform (more on this below) to host a Digital Twin of the Sydney Olympic Park with a focus on urban cooling and plant health. The Digital Twin operates 24 hours, 7 days a week taking near real-time observations from the sensor network and forecasts from the distributed Eratos catalogue; biophysical modelling and machine learning is then applied to issue smart irrigation commands to the park irrigation system. The solution runs completely on the Eratos Senaps platform, a cloud-based IoT and Analytics platform backed by AWS services.

The Eratos Platform utilises Amazon Web Services (AWS) to leverage best in class cloud scaling architecture, including Amazon Elastic Container Service (Amazon ECS), Amazon Elastic Kubernetes Service (AWS EKS), Amazon Elastic File System (Amazon EFS) and Amazon Simple Storage Service (S3).

Results and Benefits

Smart irrigation system, optimised for current conditions to maximise park cooling
Plug-and-play access to disparate, federated data resulting in rapid deployment of machine learning modelling and solution prototyping.
Standardised data and modelling resulting in easy replication to other locations.