The real R&D problem: data chaos in the lab
Modern analytical instruments and automated processes are well established in R&D labs. Yet formulation development often remains slower than expected. The reason rarely lies in the technology, but in the way data is organised. In many development departments, data is fragmented across different systems and difficult to bring together:
- Trial data lies in Excel files
- Project data is in the ELN
- Historical results sit in isolated systems
Many lab instruments store measurement data in proprietary file formats. Structured metadata about the experiment, such as the formulation, sample handling or test conditions, is often missing too. That leaves out the context that is essential for using the data later on. As a result, measurement values can only be compared, interpreted or used for data-based decisions to a limited extent.
The real problem, therefore, is not a lack of data, but the absence of structured lab data management.
Typical consequences include:
- Experiments get run twice
- Teams spend a lot of time searching for data
- Decisions are made on incomplete information
What data is generated in adhesive development?
Adhesive development generates a wide range of data types.
These include, among others:
- Formulation data such as mixing ratios and raw material combinations
- Process parameters such as mixing sequence, dispersion conditions, temperature profiles or curing conditions
- Measurement data such as viscosity, rheological behaviour, tensile or shear strength of bonded joints, ageing resistance, glass transition temperature or thermal stability
A further challenge in adhesive development is raw material variability. Properties of resins, hardeners, additives or fillers can differ significantly between batches or suppliers.
This variability affects formulation performance. Structured information on raw material batches, suppliers and trial results helps to understand these effects better and design more robust formulations.
Another important factor in data-driven material development is the reproducibility of material tests. In practice, measurement results can vary due to differences in sample preparation, test conditions or instrument settings.
For data-based analysis, it is therefore essential that trial data is documented consistently, for example regarding sample preparation, sample dimensions, curing conditions or test methods. Only when this context information is available can material properties be reliably compared and used for data-driven analyses.
R&D data is spread across different formats and systems. Without a clear structure, it is difficult to link and remains only partly usable for cross-project analysis.
Excel: not a sustainable approach to lab data management
In many labs, Excel is a central tool for data analysis. That is understandable: the software is flexible, easily accessible and familiar to almost everyone.
But this is exactly where a structural problem arises: Excel was never built for systematic R&D data management.
When teams try to manage complex formulation projects using Excel, the result is often:
- Inconsistent datasets
- Trial histories that are hard to follow
- Missing links between raw materials, process parameters and test results
- A high time cost for searching and analysing data
Excel quickly reaches its limits in R&D. Structured data management therefore calls for systems built specifically for these requirements.
Data-driven formulation development with Material Intelligence
Structured R&D data is becoming an increasingly decisive factor in modern material development. Material Intelligence describes an approach that makes data systematically usable and organises development processes on a data-driven basis.
Material Intelligence rests on three elements:
- Structured data
- Workflows
- AI
1. Structured data
Measurement data, formulations, raw materials and test results need to be linked together so that central development questions can be answered, for example:
- Which combinations of resins, hardeners, additives and fillers produce specific mechanical, rheological or thermal properties?
- Which process parameters influence the performance of an adhesive?
- Which trial variants have already been tested?
Structured data is also a prerequisite for using machine learning. Trial data needs to be documented consistently and traceably so that algorithms can identify relationships between formulation parameters, process conditions and material properties.
In material development in particular, data quality is therefore more important than sheer data volume.
Without structured relationships between raw materials, parameters and test results, the available data is of little use for machine learning and AI applications.
2. Workflows for real lab practice
Digital solutions for formulation development need to reflect the reality of lab work. This includes, among other things:
- Integration of lab instruments for automated data capture
- Structured project documentation
- Consistent trial data
- Automated reporting
A modern data platform for research and development can connect different data sources with one another, reducing data silos.
The reality in many research labs
In practice, lab environments in many companies have grown organically over time. Different instruments, software solutions and documentation methods often exist side by side.
Full integration of all data sources is therefore usually not possible right away. Many companies start instead with incremental improvements to lab data management, for example through structured trial documentation or by connecting individual instruments.
Typical data sources include, for example:
- Lab instruments such as rheometers or tensile testing machines
- Electronic lab notebooks (ELN)
- Raw material and material databases
Automated data transfer allows measurement values to be linked directly with trial parameters and formulation data. This reduces manual data entry and minimises errors.
3. AI in the lab
Adhesive development often involves very large formulation spaces. Even small changes to resins, hardeners, additives or fillers can significantly affect a formulation's mechanical, rheological or thermal properties.
Evaluating these relationships systematically requires a structured data foundation. On this basis, AI can also be put to meaningful use in the lab. Two central applications can be distinguished here: digital assistants and machine learning.
Digital assistants can support developers by:
- Searching measurement data faster
- Comparing trial results
- Identifying raw material alternatives by linking formulation data, raw material databases and historical trial results
Machine learning models can analyse historical development data and identify relationships between:
- Raw material combinations
- Process parameters
- Material properties
Such models support the identification of suitable formulation variants and the quantitative prediction of material properties based on the underlying formulation parameters. Their goal is not to replace the experience of developers. Rather, they help experts make better use of existing data and access relevant knowledge more quickly.
Why efficient trial planning is essential
Experiments in material development are often time-consuming and costly. Every new formulation requires several successive steps, from production through sample preparation to material testing.
When historical data is structured and analysable, it can be used for data-driven planning of new experiments, reducing unnecessary trial iterations.
The role of historical trial data
A structured data platform can help make such historical datasets usable again, creating an important foundation for data-driven material development.
How a data strategy speeds up formulation development
Companies that bring their R&D data together centrally through systematic data management often see measurable effects, for example:
- Fewer duplicate trials
- Faster data searches
- Better comparability between formulations
- Data-based decisions
Why lab digitalisation is becoming critical now
Requirements placed on materials continue to increase. New products are expected to perform better, be more sustainable and reach the market faster.
Sustainability requirements are also significantly increasing the complexity of modern material development.
Companies need to evaluate alternative raw materials, reduce the use of substances of concern and account for new regulatory requirements. In adhesive development, this often means adapting existing formulations or developing entirely new ones.
A structured data foundation can make these tasks in formulation development considerably easier. Historical trial data helps, for example, to identify suitable raw material alternatives faster or to better estimate the impact of new components on material properties.
This makes data-based formulation development an important factor in developing more sustainable materials as well.
Given these rising requirements, a clear data strategy in R&D is developing into a competitive advantage. Companies that invest in structured lab data management and data-driven development are building the foundation for more efficient, future-ready adhesive development.