A Proof of Concept and feasibility study utilising space technologies to advance the aquaculture markets in Western Australia remote and regional areas.

There is enormous opportunity to grow the market of the aquaculture sector in Western Australia, yet there are several challenges that need to be addressed for this to happen.

To grow the sector, producers need to access data that can inform them of suitable shellfish growing sites and use this site characterisation to better assess stock carrying capacity and feasibility of farming. With this intelligence, producers can target suitable areas to develop with greatly reduced risk. This pilot project will prototype a software solution ingesting EO, in-situ and model data, which can be trailed by our end-users to assess whether such a tool could assist them in determining the suitability of future sites. The prototype will focus on two sites, one in the Pilbara and one in the Kimberley, both remote areas of North West Australia. This solution will be designed so that it can also evolve towards an EO-based farm site monitoring tool.

Australia is in a privileged geographical position when it comes to Earth Observing space missions: our land is representative of almost all climate zones in the planet, terrestrial and aquatic systems and covers most biome types.

Increasingly, multiple stakeholders – including industry, government, defence, academia and NGOs – rely on EO data for their mapping, monitoring, and modelling activities across several application areas. For EO data and products to become widely used and adopted, providing the right level of “trust” is critical, especially on key sectors where benefits of EO could be invaluable such as agriculture, mining, emergency services, and natural resources in aquatic and terrestrial environments. This trust comes from understanding the explicit link between the satellite derived information and sampled ground measurements (calibration) as well as knowing its level of accuracy (validation).

The project aims to identify promising areas of research in the field of pre-operational testing of Earth Observation sensors (calibration), and algorithms and analytics (validation) for Earth Observation. It will concentrate on developing a prototype testbed for calibration and validation for commercial or research satellites.

Australian industries and land managers are required to restore and rehabilitate land that they have disturbed. However, these industries and government regulators lack Earth Observation (EO) tools to effectively characterise progressive rehabilitation of these disturbance footprints.

This project aims to engage the mining industry and land management agencies to develop a clear understanding of EO tools needed to improve government monitoring, and industry management approaches to mine rehabilitation and restoration at the property scale (including stochastic event monitoring). We know from recent discussions with multiple mining companies (e.g. BHP, Anglo, and Rio Tinto) that there is a compelling argument to link remotely sensed observations with rehabilitation success criteria.

This project includes a needs analysis based on industry and government engagement; a literature review on EO for rehabilitation monitoring and tool development; and the development of a roadmap that highlights current capabilities and recommendations for future research.

EO4MDA is ultimately planned to be a multi-phased project aimed at enhancing Earth Observation practices and to generate robust and reliable information about the maritime domain to support the Commonwealth interdict and enforce laws across its Exclusive Economic Zone and coastal areas.

EO4MDA will produce a downstream processing capability to correlate, federate and reason from data sourced through ground, air, and naval-based systems integrated with space-based Synthetic Aperture RADAR (SAR) data reports generated by the COSMO-SkyMed constellation data, satellite-based optical data as well as satellite-based AIS and SIGINT collections.

This project, EO4MDA Phase 1, is the first collaborative step to test the limits and bounds of new AI/ML/statistics-based processing methods in a realistic scenario employing constrained space-based Earth Observation data. The aim is to resolve operational queries (anomalous vessels and vessel behaviour) put to system in a real world demonstration with maritime decision-makers with space-based Earth Observation.

In particular, it is currently best practice to exploiting non-cooperative characteristics of vessels from space imaging sensors for the detection of anomalies:

• vessels with failed or malfunctioning GPS and/or transmitting equipment;
• vessels that deliberately “turn off” the System so as to avoid detection;
• vessels that, because of their smaller size, are not under the obligation of having an on board positioning system as well as sport fishing vessels;
• abnormal behaviours

This 15-month project seeks to  develop a commercial assessment of a focused selection of Earth Observation (EO) end user needs to define future SmartSat CRC Research activities and projects.

The research will address what end user problems EO can address and which of these problems people will pay to have solved. The output will be a report covering defined problems, customers, commercial viability, due diligence and top level mapping to SmartSat partner capabilities that can inform research that can be progressed now or requires capability development to progress.

For industry to develop new requirements for sensors, it is imperative that industry users are able to understand and demonstrate the specifications of the proposed sensors in relation to their specific use cases. At the current time this is achieved by utilising parameters that have been used in airborne collections or by estimating the requirements from expectations of the use case. This includes, number of spectral bands, width and positioning of bands, spectrum coverage (eg 380 to 1000 nm for aquatic indices; 400-2400nm for vegetation indices, 1000-3000nm for mineral detection), as well as radiometric and spatial resolution. Having a tool that can simulate these requirements and test the sensor trade-off outcomes, provides a capability that will enhance the efficient and cost-effective design of future sensors, tailored to industry requirements as well as empower the value-adding information industry (VAI) to develop appropriate Earth observation algorithms.

A virtual testbed in the form of an end-to-end simulation tool is an essential component of the design process of any new Earth imaging system (CEOS, 2018). These tools in Australia currently consist of ad-hoc compilations of different advanced software packages and their parameterisations spread over research groups. Optimization of sensors requires understanding the effects of instrumental and environmental parameters on the resulting image characteristics, which can be achieved through precise simulation of sensor images. For this purpose, we propose to develop a SmartSat CRC simulation suite for Earth Observation Sensor Design as well as VAI algorithm development and testing. The software enables simulation of satellite EO sensor measurements over most possible variations of aquatic ecosystems, using a range of sensor design specifications, and subsequent simulation of at-earth-surface image products. There are no commercial implementations of this capability as a software suite. The different stages of the processing chain for SmartSat CRC Earth observation (from high altitude platform or satellite) processing chain are depicted below. They comprise a scene simulator and its inversion process which are briefly described in the following.