The success of Modern Manufacturing Needs Strong Foundations in Good Governance

The success of Modern Manufacturing Needs Strong Foundations in Good Governance written by Meighan Heard, Executive Director, Commercial and Corporate, Mitsubishi Electric Australia

In 2018, a research paper from MIT and Stanford University concluded that three commercially released facial-analysis programs from major technology companies demonstrated both skin-type and gender biases.  The research received high profile coverage after reporting results demonstrating that, for each of the programs assessed, the error rate for light-skinned men was never worse than 0.8% but for darker skinned women, it was more than 20% in one case and more than 34% in the other two. 

The issue was discovered by one of the authors who, while working on a project as a university graduate student, discovered that the facial recognition technology used in the project didn’t work reliably with darker-skinned users. She could only get it to work if she wore a white mask. This meant the team had to use lighter-skinned team members to demonstrate it for the purpose of the project. She went on to submit photos of herself to a number of commercial facial recognition programs which, in many cases failed to recognise the photos as featuring a human face and, if they did, consistently misclassified her gender.

Upon investigation, it became apparent that the technology failed to work properly because the systems simply weren’t familiar with faces like hers.  In fact, the data set used by the relevant companies to assess performance was found to be significantly biased towards males and people who were white.  This meant the algorithisms did not perform as intended once put into operation.

It is often assumed that machines are neutral.  This research confirmed they are not. They continue to be influenced by decisions taken during their design, development and deployment. 

This research project highlights the importance of ensuring the right parameters are put into place when developing a new solution using technologies such as artificial intelligence and machine learning.  A good deal of effort goes into defining, testing, refining and perfecting these parameters and it is critical to get this right.  It also highlights the importance of being aware of ethical and legal issues associated with implementing new technology, and shines a light on the need for a strong risk management framework to assess and evaluate new digital initiatives before they are rolled out.

It is absolutely critical for all manufacturers to have the right structures, systems, policies and processes in place to ensure their digital transformation initiatives will be a success. This is where good governance comes into the equation.  Governance is the system by which an organisation is controlled and operates, and the mechanisms by which it, and its people, are held to account. It is the foundation on which good decisions are made and it is relevant to all organisations regardless of size, shape or type.

As the digital revolution continues to be deployed at faster and faster pace, it is critical for manufacturers to ensure they have the right governance structures in place to support their digital initiatives.  These initiatives can fundamentally change the business model, culture, workforce and operations of the organisation. 

In the most recent version of KPMG’s Directors Toolkit, companies are encouraged to ask questions such as the following when considering the implementation of automation and artificial intelligence:

(a) what is the scope of the impact on the organisation and its industry, and the cost of missed opportunities;

(b) how well do we understand where automation and artificial intelligence can deliver cost efficiencies;

(c) how well do we understand the impacts on the customer experience and on driving growth and customer loyalty;

(d) are we challenging and rethinking existing ways of working or simply automating existing processes and approaches;

(e) what are the ethical issues we need to consider as we adopt greater automation and artificial intelligence into the business; and

(f)  do we have a robust risk management process and control framework to manage automation and artificial intelligence.  

In unpacking these questions, it will be critical for organisations to have good data and analytics to support decision making.  This is essential for good governance in any organisation. 

Mitsubishi Electric provides technology, systems and software to help companies collect data, analyse that data and use that data to obtain helpful insights into their operations and activities.  At an enterprise level, the products and solutions provided by ICONICS empowers companies to leverage their information so as to achieve immediate benefits for any application.

The ICONICS solutions include rich, dynamic dashboards convey concise, role-based information for any manufacturing, industrial or building management company. The information collected can be exported and manipulated by users to obtain additional insights as required. It can also be used to support OEE and KPI analysis, statistic process control and predictive/preventative maintenance needs. The solutions are stackable and can be rolled out on both a small scale – for organisations who are just starting their digitalisation journey or only need a simple solution – and on much larger scales for complex and multi-dimensional organisations that may operate across multiple sites.

As the digital evolution in across the Australian manufacturing industry continues, all companies will benefit from the implementation of a solution such as ICONICS.  Mitsubishi Electric is proud to support manufacturers in establishing and maintaining the foundations of good governance through collection and delivery of meaningful, useful and accessible data to support strong decision making in this regard.

Further reading:  

Mitsubishi Electric Australia will be exhibiting at the Modern Manufacturing Expo on 20-21 September


Leave a Reply

Your email address will not be published. Required fields are marked *