Wouldn’t it be great if you could predict when a critical asset was about to fail, before it failed? Because when big assets fail, the results can be catastrophic. The deadly Californian bushfires of 2019 were caused by a defective power asset that did not get detected. The result:

• 86 people died

• $30 billion in litigation

• A $23 billion fine to escape bankruptcy.

Since assets like power poles, transmission towers, railway tracks or roads are often distributed across a vast area, they're difficult to access and there are just too many of them to inspect.

As a result, they're difficult to inspect and the capture and review process for photos of each asset can take months. A dangerous defect can get missed in the interim and cause a catastrophic outcome.

Could Machine Learning be used in asset management to prevent disasters like this from happening again? We think it can.

So, what if we could replace the manual assessment of those photos with machine learning assessment so that you could assess more photos of your assets more quickly and:

• Instantly eliminate the vast majority of photos that don’t have defects

• Triage the tiny percentage of photos that do

• Present those to your expert team for review and get those assets repaired quickly

Why isn’t machine learning already being used in asset assessment? Well, they would if they could, but they can’t. In order to train the computer to spot a defect in a photo, the computer needs to be shown millions and millions of photos of that defect in order to know what the defect looks like.

But who has the millions and millions of photos of the defect in order to train that computer? We do.

We’ve developed a very exciting piece of software that can generate the millions and millions of photos needed to train the computer to spot the defects.

We provide this missing piece of the machine learning puzzle that makes machine learning accessible and affordable for all. For example:

The standard industry cost for a human to assess one photo is $20 per photo. We can assess it for 10 cents. To manually assess 2 million photos and complete one full assessment cycle costs them $40m. We can do it for $200,000. What’s more, it takes up to 9 months to capture and review all those photos. We can do it in a week.

So now, with our software we can:

• Increase the frequency of assessment

• Reduce the cost of compliance

• Reduce the risk of defects causing disasters

1. Software licencing fee

2. SaaS platform

Once several broader models have been created we will create a retail offering on the SaaS platform allowing self-signup to upload photos and have them assessed by one or more models.

We are working with intermediaries like engineering firms who are pre-qualified customers for our technology.

We are also running events with Melbourne City Council and partnering with AWS to educate the market and fill the sales pipeline with qualified leads.

We have 3 pilots in the opportunity pipeline, one with AusNet Services, and we are on the cusp of signing a contract with a global engineering consulting firm to detect cracks in concrete roads.

The co founders have been:

• Winners of the Bendigo Bank Business Award for Excellence

• Nominated in the 2020 Top 50 Small Business Leaders Award

• Winners of the Asset Management National Conference Pitch Competition

• Finalists with the Amazon Web Services (AWS) Machine Learning Pitch Event

• Nominees in the Top 20 Hottest AI Companies in Australia.

CEO: Bernadette Schwerdt is a marketing strategist with 30 years’ experience in the sector. She has worked on global brands such as Apple, American Express, Colgate, BHP, McDonalds and many others and created award winning campaigns for a raft of others.

She is an experienced entrepreneur and the director of a number of businesses including the Australian School of Copywriting and her consulting firm, BMS Consulting. Her clients include Lend Lease, Stockland, the Property Council of Australia, Mirvac, KPMG, Goldman Sachs and many others.

She is the author of two bestselling novels on digital entrepreneurship (How to Build an Online Business and Secrets of Online Entrepreneurs.

CTO - Toby Tremayne is a software engineer, coder and hacker and has been developing cross-disciplinary enterprise software in a many industries for over 22 years. Working with emerging technologies like Machine Learning, Blockchain, Web, Mobile, Cryptographic Identity, Encryption, VR/AR/MR & Game Development, he is a big picture architect who combines different disciplines and technologies to create something new.

Our automation and reporting platform allows B2B and retail SaaS for object detection on individual models or industry wide "commodity" models like Building Maintenance, Power Equipment, Rail etc.

We aim to make machine learning accessible for all, putting simple tools in the hands of non-technical users and aiding the proliferation and development of Machine Learning as a discipline and a customer tool.

If you are an asset manager, engineer or you are responsible for maintaining assets, and would like to do it more quickly, safely and affordably, book a demonstration of our software or visit www.machinedreams.com.au to download a free copy of our report about how you can apply machine learning in your business today.

Contact Bernadette Schwerdt at:

E: bernadette@machindreams.com.au

M: 0419891932