The market for new inks for coding and marking and digital printing applications is always changing – with new formulations constantly required to keep up with regulations and customer demands. Unfortunately, developing new ink formulations can be a lengthy, often labour-intensive process.
Traditionally, a reliable formulation, that meets the needs of the end-user and adheres to all necessary safety and regulatory requirements, has been the result of extensive collaboration between chemists, technical engineers, and field service personnel running customer trials. It can take months, if not years, to perfect.
In the following article, Josie Harries, Group Programme Director, Domino Printing Sciences (Domino), sheds some light on this subject and demonstrates how Domino is embracing digitalisation to help speed up the research and development (R&D) of new ink formulations for its coding and marking, and digital printing customers.
Digitalising R&D
At Domino, we recognise that achieving optimal print quality, and minimal downtime are of key importance to our customers, which is why we develop and manufacture the majority of our inks and fluids in-house to ensure exceptional quality performance with our equipment. In the UK alone, this equates to the production of roughly 3 million litres of ink each year, from our specialist inks facility in Liverpool.
We also recognise the fact that the world is changing – and today people expect businesses to do more with less, and to be able to create the same exceptional quality products, in a fraction of the time. So, we asked ourselves: how can we speed up our in-house ink development, while still ensuring high-quality, high-performance products that will fulfil our customers’ needs?
To crack this conundrum, we have started on a journey towards digitalising our R&D efforts, by utilising automation, predictive modelling, and machine learning to improve our processes, speed up laborious tasks, and ‘Do more’ with less.
Automating the process
In the labs, our expert chemists typically spent a lot of time on low-value tasks such as weighing out materials, mixing formulations, and testing samples – these simple steps were identified as a potential area where digital technologies could be used to increase productivity.
Domino invested in a robotic platform capable of handling chemical powders and liquids, and mixing sample inks based on predetermined formulations, with minimal human intervention. Automating this part of the formulation process resulted in less than half the effort normally required to produce over 1,300 samples – effectively achieving a year’s worth of sampling in just six months. In addition, we have recorded a 57% reduction in the amount of chemicals used during experimentation, this figure will decrease further as we move more development tasks onto the robotic platform.
Overall, the time and cost savings effectively allowed us to pay back the cost of our new equipment within six months. The precision of new automated equipment has also enabled us to reduce the size of each sample from 100g to just 8-10g. Fewer materials equal less waste and cost, and, as the equipment washes itself, less manpower and fewer solvents are used – the cleaning function has resulted in an 80% reduction in solvents, a saving of 330 litres in 2020. This results in a win-win from a productivity, cost, and an environmental perspective.
Predictive modelling to accelerate learning
Another area identified as potentially benefitting from digitalisation was our data analysis. We set out to explore how we could use experimental data to create statistically designed experiments for ink formulation – that is, designing experiments using statistical software to identify sample ink recipes with specific properties.
To make the best use of the data our chemists generate during their R&D lab work, we have developed a bespoke electronic lab notebook. This allows the team to efficiently plan, record, and access key experimental data. As it is bespoke to our R&D work flows, it easily integrates with the day-to-day lab practices. We have also introduced functionality that allows the teams to see the chemical stock levels in our chemical stores to ensure that all materials are available before work starts, therefore reducing wasted time. We also use the system to perform and store our health and safety assessments, the outcome of which is automatically printed on the sample labels used.
We use this data as a basis for predictive modelling to identify formulation candidates, which can significantly speed up the sampling process. The idea is to utilise the wealth of data created via experimentation to explore the design space more effectively and come up with a better starting point for proof-of-concept designs.
Virtual formulation using machine learning
Within this new approach, we have also explored using data engineering to support machine learning to target values for ink formulations. If we are looking to produce an ink with a certain set of properties, for example quick dry time on flexible film in a cold environment, we can feed this information into a machine learning system to identify a starting hypothesis for an ink design. The system can also help us to identify where there are weaknesses in our data and suggest experiments to improve this.
Conducting experiments utilising computers – also known as in-silico experiments – is helping us to improve the quality of our data and speed up the time taken to produce new formulations in the event of chemical reclassification. This means our chemists are also able to spend more time on novel chemistries research, which has greater customer benefits down the line.
From a sustainability perspective, in-silico experimentation also pays off in terms of waste reduction. When carrying out virtual reformulation, the starting point for ink designs takes place on a computer before doing any lab work, reducing the overall quantity of chemical waste created as part of the design process.
The findings obtained via automated formulation and testing enrich the design data, which in turn allows for more precise in-silico experiments. This cycle of innovation will continue to drive our research and allow us to deliver new reliable ink formulations in response to changing legislation and customer demand at unprecedented speed.
Future efforts
Of course, there is still work to be done – and as such we are constantly evaluating our processes and inspiring our chemists to achieve more within their work.
It is an ongoing cycle, with new process improvement creating more time for R&D teams to drive scientific and manufacturing innovation forward. Our chemists are now better able to take advantage of new training opportunities and spend time researching and experimenting with areas outside of their immediate field – such as data management and coding.
We have also launched several internal forums and working groups to facilitate research and collaboration across the R&D community. These groups provide a space to work through problems, share knowledge, and spawn ideas for future projects, feeding into an ongoing cycle of continuous improvement.
In addition to this, we also look outside of the business, at how we can work with others in the wider scientific community to foster innovation. A team from Domino was recently involved in a project with Cambridge University’s Institute for Manufacturing, to investigate a historic issue with ink build-up on continuous inkjet (CIJ) printers. The findings from the project will feed into our future predictive modelling and machine learning activities to help to facilitate the design of more reliable CIJ ink formulations.
Conclusion
The technologies described here may sound futuristic – but really, we are only just scratching the surface of what is possible. Scientific innovation is now about so much more than chemicals and test tubes – today, chemists working in all industries will find themselves well placed to begin exploring how technological skills, in areas such as coding and data analytics, could help drive innovation in their field.
At Domino, we have a vision of the ‘chemist of the future’ – a tech-savvy, digitally-minded individual, comfortable with both the chemical and technological – whose role will be integral to our R&D efforts. We see a bold future for how digital processes will be used as part of ink formulation processes, and the benefits that this will have in terms of speeding up time to market for new formulations, and making gains in sustainability and waste reduction – this is really just the beginning.