Jonathan Legault

Managing Director at Corematic

The Australian Institute for Machine Learning – who are pushing to establish a National Centre of Excellence in Machine Learning – found that Australia is spending just a fraction of its GDP on artificial intelligence and machine learning when compared to its international rivals, and will lose jobs without increased investment in the technology.

While the US and China are leading the way in AI investment; Australia’s investment in AI is well behind comparable nations including South Korea, Singapore, and Japan. Australia was the first country to automate its ports and mine sites, both industries important parts of the nation’s history, and in many cases their culture! AI experts, engineers, and data scientists all agree that we have been presented with a huge opportunity to build on our innovative past and create a new foundation with Machine Vision and AI that will benefit the country and future generations.

There’s a call for Machine Vision in Australia

Now that it’s coming into its own, companies everywhere are exploring the benefits that Machine Vision brings, especially in automotive, machinery and agricultural engineering sectors. When it comes to innovation, machine vision is one of the most technically advanced solutions for nations that seek to increase productivity and national income while developing new products and services capable of creating new business models, jobs, and opportunities. With the overall production growing at a steady rate, the demand for machine vision is expected to be influenced over the coming years, supporting the market growth.

New opportunities for the use of Machine Vision will be everywhere, and this is something Australia cannot escape! In reality – it’s already started: CSIRO’s Data61 has developed drones capable of travelling in GPS-denied environments utilising 3D LiDAR technology – they are able to travel down mine shafts to safely inspect hard to access areas of underground mines (so people don’t have to), while mapping along the way. The New-Zealander firm Vector has created a data-driven energy supply platform in Australia and New Zealand that partners expect to enable tailored products for customers who are increasingly turning to solar power, batteries, and electric vehicles.

Overseas, a Harvard University group has spent 12 years creating a ‘robotic bee’ capable of partially untethered flight thanks to artificial muscles beating the wings 120 times a second. The ultimate goal of this project was to create a swarm for use in natural disasters for artificial pollination given the devastating effects of colony collapse disorder on bee populations and consequently food pollination.

Machine vision is going to be everywhere before long

The major advantages of MV are to increase productivity and provide high flexibility in the production process by improving the quality control inspection phase, and increasing accuracy and speed in automated material handling.

This can directly benefit manufacturing firms and industries that are starting to understand the potential of MV systems, particularly where redundant or mundane tasks such as inspection, need to be performed with seamless precision.

So, how will you be sure that your business capitalises on this opportunity to be a first-mover, and maximises the benefits?

Machine vision isn’t just about the overall comprehension of the system and how those critical components interact with each other – Machine Vision is also about the fine details, the expertise of each component in the system interacting together to work reliably and generate repeatable results.

Machine vision is a set of technologies that gives machines greater awareness of their surroundings. It facilitates higher-order image recognition and decision-making based on that awareness. Machine vision makes sensors throughout the IoT even more powerful and useful. Instead of providing raw data, sensors deliver a level of interpretation and abstraction that can be used in decision-making or further automation. Machine vision can be used with sensors, robots, and other IoT technologies. Just like another tech, machine vision can free up valuable employee time by performing repetitive, time-consuming tasks. 

And that’s why, at Corematic, we’ve built a team of passionate engineers capable of collaborating with companies and providing an interdisciplinary approach to R&D and consulting.

We have a thorough awareness of entire vision systems – Imagine a robot that hits a button as soon as it is prompted… It might sound simple enough, but the entire process of making the robot move its arm at the right time, to the right place, with the right force, is far more complex than simply a ‘Visions System’.

Our expertise involves complex systems that enable environmental features to be identified by hardware using different types of information from RGB, infrared, and spectral cameras.

What does it mean?

It means that once this vision is available, the possibilities for use are endless. Whereas human inspectors once had to touch and individually verify each workpiece as it came off the line, automated inspection stations have changed everything about the process. Now, we can utilise information that other systems cannot produce, and utilise robots for performing tasks that we cannot due to limitations such as size, distance, danger, or the mundane. We can remove certain human elements and the risk of inconsistencies, failures, or hazards that come along with them.

We are using Machine Vision to help businesses prevent parts damage and eliminate the maintenance time and costs associated, benefiting our customers in wide range of industries from agriculture to biopharma; smelting to construction. One of our latest technologies, TallyOp, has combined vision systems, sensing, and sensors to manage risks and increase productivity. This technology allows producers to not only identify defects during harvest but also generate a heat map to identify the best performing areas of a crop. This heat map becomes useful for soil testing in order to better replicate conditions, and make better-informed irrigation and fertiliser choices, while defects can be identified and sorted at a far greater (and more accurate) rate than would be possible by hand. This technology gives farmers the information they need to better manage field health and have access to data in real-time for faster decision making and input, resulting in improved product quality, higher yields, and lower production costs.