Research projects under the CSIRO’s Future Science Platforms are paving the way for new breakthroughs across agriculture and food production and the manufacturers supplying the sector.

CSIRO’s research is broad and its research capabilities traverse many scientific and engineering disciplines as it seeks to be a trusted adviser to Australian industry. A strong focus of the CSIRO strategy is to aid the transition of Australian businesses towards more advanced, sustainable and profitable practices – whether that be in agriculture, mining, energy or manufacturing.

The recent announcement of CSIRO’s Future Science Platforms (FSPs), an investment in breakthrough research growing to $52m per year over five years, demonstrates the organisation’s commitment to undertaking fundamental science in areas that will be critical to the opportunities and challenges we face in the future. One Future Science Platform is Active Integrated Matter (AIM). The AIM FSP, led by CSIRO’s Dr Danielle Kennedy, is looking towards 2030 with a focus on new discoveries that occur when you combine what are core research strengths for CSIRO: materials, processing, sensing, robotics, autonomous science and big data.

According to a number of future forecasts, there will be dramatic and wide-ranging changes in the world by 2030. Personalised food and therapeutics, increased regularity of extreme weather events, and human-free autonomous manufacturing and maintenance environments known as dark factories are just some examples of what we are to expect. AIM’s goal is to provide Australian industry with the tools needed to capitalise on these predicted changes and ensure preparedness for future population, climate and environmental scenarios.

Within AIM there are several focus areas called test-beds: Eliminating Food Loss; Morphing Equipment; Autonomous Science for Physical Construction; and Smart On-Farm Manufacturing.

Smart On-Farm Manufacturing

The Smart On-Farm Manufacturing test-bed is set to place Australia ahead of the forecast disruption in current agricultural supply and value chains. CSIRO is investigating the introduction of precision bespoke chemical manufacture, delivered by automated mechanisms and supported by environmental data.

Take herbicides, for example. Currently, herbicides are applied to fields at 1,000 to 10,000 times the optimal dose, posing waste and cost challenges for farmers. These issues could be addressed through the production of “customised” formulations – on the farm, as needed, and to the precise formulation as required by crops at any particular time. Moreover, these tailored formulations could be produced and delivered with a granularity to the resolution of centimetres.

This dramatic change in manufacture and delivery of agricultural products needs to be informed and enabled through the development of a range of sensors detecting variants such as temperature, rain, humidity, water hardness and chemical composition. Such sensors would be mounted on drones or small robots for crop growth monitoring. The collected data would inform an automated system that could then recommend and distribute fertiliser at an appropriate level while monitoring run off to optimise future applications.

A smart on-farm agricultural system would enable the production of pesticide, herbicide and/or controlled-release nutrients at the farm by shifting precision chemical manufacture and formulation to the farm gate or centralised agri-hubs.

Morphing Equipment

The world is experiencing a Cambrian-like explosion in robotics. The successes of robots in specific applications are underpinned by breakthroughs in affordable computation, advanced materials and manufacturing processes as well as the availability of high-bandwidth communication. AIM’s Morphing Machines test-bed is pushing the boundaries further by developing robots that can adapt their shape and properties to suit environmental and task requirements.

Imagine a robot that is deployed by a first responder in a disaster area after a building collapse. A morphing robot would be able to change its body size and shape to squeeze through tight spaces in rubble to search for survivors trapped underneath and provide vital resources like glucose and water for sustenance during the rescue process. Or imagine the dangerous task of confined space inspection of hazardous environments like fuel tanks, pipelines or sewage systems. A morphing robot would be able to adapt its structure to effortlessly navigate these complex environments, keeping infrastructure inspectors and maintenance workers out of harm’s way.

The ability to change a robot’s structure and properties on demand has the potential not only to  significantly improve the efficiency of current industrial inspection processes, like for example in confined space inspections, but also to develop new tools for rapid search-and-rescue disaster response and natural habitat monitoring.

Scientists across CSIRO are working together under the AIM umbrella to develop novel materials and processes that enable design and deployment of the next generation of flexible, lightweight morphing field robots.

Eliminating Food Loss

Approximately 795m people around the world are currently undernourished. By 2050 forecasts predict 9bn people will need to be fed in an environmentally and nutritionally sustainable way. Food and nutrition security poses both a major global challenge and an economic opportunity. Roughly one third of the food we produce is lost in the supply chain between production and the consumer. This food loss is contributing to food insecurity and negative environmental impacts, with limited strategies existing for the recovery and reuse of food loss for production of value-added products.

For manufacturers there are many opportunities to increase the efficiency of food processing and reduce food loss along the value chain, from production to consumption, while maintaining nutritious food products with high consumer appeal. Over previous decades the focus has been on efficiencies in production and manufacturing processes, more efficient supply chains and fresh produce storage methodologies.

CSIRO proposes that the next step in food production efficiency is food loss avoidance, by utilising the edible portion of lost or wasted produce both on the farm and in the supply chain before it reaches the consumer. Reducing loss could be using the offcuts of vegetable crops. It could be developing new methods for processing non-saleable produce or catching produce at peak nutrition. Addressing food loss offers new prospects for food production and nutraceutical markets with potential economic, environmental and social gains.

The goal of AIM’s Eliminating Food Loss test-bed is that producers and manufacturers will be able to make informed decisions on the suitability of lost food to re-enter the food chain. The focus will be on producing food ingredients, supplements and food products that are safe, nutritious and sustainable. Mapping where food is lost along various supply chains in Australia will be an important prerequisite for intelligent decision making. This will be coupled with advanced sensing technologies, which enable decisions to be made on the fitness-for-purpose applications of food loss and waste. The further development of efficient processing and preservation technologies will allow the stabilisation and conversion of underutilised edible biomass.

Autonomous Science for Physical Construction

As the rise of the machines across industry continues unabated, the limitations of those machines are becoming increasingly exposed. Hardware issues can include component breakage, maintenance, energy supply, and physical limitations such as servo angular ranges and torque limitations. On the software side, control algorithms can lack adaptivity and fail to respond effectively to dynamic scenarios.

Combined, these hardware and software factors shackle the potential of robotics in industry, and limit the number of feasible use cases by requiring persistent human intervention. AIM is solving some of these problems by pushing the limits of intelligence algorithms and construction techniques. This work seeks to close the loop on human intervention.

AIM’s Autonomous Science for Physical Construction test-bed will focus on ways to improve the construction of new robotic hardware without human involvement. To create new hardware, the team is applying advanced machine learning techniques to automatically generate fit-for-function robot hardware, which can then be tailored to various needs: capability, cost reduction, and long lifetime. Because components are created specifically for a given operating environment, they can outperform off-the-shelf solutions in their specific niche. Also being investigated is the use of flexible manufacturing techniques, allowing capable, bespoke robots to be created in one shot.

The team is concurrently developing intelligence algorithms that will allow robotic systems to understand changes in their environment, and then respond and adapt without the need for human intervention. This means that robotic systems can observe parts of their environment, and test hypotheses based on these observations. They can then use these results to improve their own performance, and also share findings amongst themselves in a form of whole-process optimisation. The algorithms will also allow robots to adapt to hardware degradation.

By endowing robots with intelligence, and enhancing their capabilities, the team aims to increase the pervasiveness of robotics in industry and deploy them into challenging areas to meet the demands of Industry 4.0 and beyond.

AIM represents an opportunity to influence the modernisation of the materials, manufacturing, processing, maintenance and environmental management industries through the broad adoption of cost-saving autonomous and robotics technologies and advanced materials processes – all underpinned by big data and intelligent algorithms. Scientists from across CSIRO’s eight multidiscipline business units are working together under the Active Integrated Matter Future Science Platform to undertake the fundamental research required. The AIM team is looking to develop partnerships between academia, research, industry and business in a variety of flexible ways, including through research collaboration, researcher placements, consultancy services and commercialisation of intellectual property (IP).