Are organisations in Australia slow to adopt Machine Learning?
Here at DiUS, we’ve been talking with many companies about leveraging ML over the past five years and found most organisations want to adopt ML, yet this transformative technology is not being adopted at the rate it should be.
In fact, our experience has been that a great proportion of organisations struggle to move beyond proof-of-concept or pilot stage. Picking the right problem, data challenges, model accuracy and application integration can be big blockers, either delaying or preventing ML project success. However, we recognise that our observations may be incomplete, so we undertook a pulse of our clients and the broader Australian market.
The survey results from the 205 respondents confirmed what we are seeing: more organisations could be driving success with ML. There is a strong appetite for ML in the Australian market with 82% of organisations interested in ML, but only 21% in production.
Other key findings in the report include:
- ML adoption is going to accelerate. 86% of respondents see ML as critical or one of several important technologies going forward, and 49% of those who have not yet started plan to do so in the next 12–24 months.
- Invest in data. Data-related challenges are either the top or second most reported challenges once the ML journey is started. The importance of data quality, data engineering and building appropriate data infrastructure and pipelines to enable ML initiatives cannot be overstated.
- Australia could be facing a ML skills gap. Only 69% of organisations with models in production report sufficient ML capability.
- Top ML use cases are internally focused… for now. The top two business areas are operational efficiency (48%) and business decision making (46%). Going forward, respondents plan a shift to both an internal and external focus: operational efficiency (57%) and customer experience (51%).
- Organisations can succeed with ML by making it a priority. 79% of respondents achieving success with ML have a strategy, suggesting that focus and investment drive outcomes.
The majority (81%) of those organisations that are in production are reporting successful business outcomes. Some examples include:
- bolttech built a new kind of customer experience. Using pioneering remote diagnostics technology, bolttech can quickly and easily onboard customers onto device protection plans. Customers simply hold their smartphone in front of a mirror and move through a sequence of tests, powered by next-gen machine learning and computer vision technology. The result is a zero-touch risk mitigation tool for bolttech and a best-in-class experience for customers. Link to case study.
- Datarock analyses digital photos to extract new value in mining. Datarock is a machine learning-powered, cloud-based drill core image analysis platform that provides accurate, fast and consistent high-resolution information about a mineral deposit’s geology. Using computer vision and image analysis, Datarock supports more efficient decision making throughout the exploration and mining process, delivering important productivity and throughput savings to a mine’s bottom line. Link to case study.
Ultimately, the pulse survey confirms there is no one formula for success; there’s no one application or area doing disproportionately better than others. There are some significant challenges, however they are not insurmountable with organisations still progressing despite these hurdles.
Here at DiUS, we remain ever optimistic, understanding that ML is still an emerging area and there is no universal ‘best practice’ yet. There’s no set and forget solution for ML. The right approach, we’ve found, is to focus on the right problems, take an experimental approach, invest continuously in the latest advancements including cloud vendor tooling and open-source frameworks, build inhouse ML and supporting capabilities and ensure organisational buy-in.
Want to know more? The Machine Learning National Pulse report is available for download on the DiUS website. It outlines each stage of the ML journey and provides some key considerations and tips for organisations to consider when pursuing ML projects.