Designing Custom Research Instruments: Lessons from Medical Device Engineering

Engineering Challenges in Biomedical Research

When I left my role as a mechanical design engineer from the product design industry to pursue PhD research in biomedical engineering, the cultural shock I found was not entirely from the science but the engineering approaches. 

In an industry setting, developing a medical device under standards, such as ISO 9001 and ISO 13485, means every design decision is precisely documented – the paper trail is fundamental. However, in a university setting, custom-designed devices are often inherited from past student projects where design workflows are often non-traceable and segmented. Instead of design history files, what is often passed on is a disjointed trail of sketches and notes siloed inside individual notebooks or buried in disconnected project folders. 

Research is inherently driven by the need to explore the unknown. To suit the highly specific and novel research objectives, researchers often need to build new bespoke devices and outcome assessment protocols. It’s a hallmark of academic innovation, but this DIY approach often introduces challenges when it comes to concept realisation. 

Specifically, take my research in cartilage tissue engineering for example. Our recent systematic review published in Advanced Healthcare Materials highlighted a notable degree of methodological variability across different research laboratories. The review identified areas where inconsistency could come from: experimental protocols, equipment variations, biological samples, and culture conditions, among others. 

Most research groups are working within very practical and tight constraints, so they have to navigate inherent biological sample variation, compromise with existing legacy equipment and balance ambitious research goals against limited funding and human resources. However, when different laboratories are utilising in-house testing rigs or highly specific protocols, it becomes increasingly difficult for the wider scientific community to reach a clear consensus on which ideas are truly the most effective for future implementation. The innovation is undoubtedly happening but establishing a unified baseline remains as a major hurdle for downstream clinical translation. 

Why Custom Research Methods Can Limit Reproducibility

Inconsistent methods, highly customised (and sometimes sparsely documented) device setups, and non-standardised testing protocols can create a bit of a grey area. How do we compare results across different research groups? For instance, if one were to evaluate the local mechanical microenvironments experienced by tissue samples in a custom-built bioreactor, the results are closely tied to the technical specifications of that exact device. If another group attempts to replicate the study using a differently configured setup, comparing those results is often like comparing apples to oranges. 

This methodological variability naturally limits research reproducibility. More broadly, it creates friction when attempting clinical translation. Regulatory bodies and commercial developers tend to prefer data with high traceability and empirical certainty. When data lack the structured rigour typically expected in industry, the pathway to commercialisation becomes incredibly complicated. 

Applying Engineering Design Controls to Biomedical Research

We can build a bridge by borrowing a few pages from the industrial Quality Management System (QMS) playbook. While implementing a full-scale ISO-certified QMS is rarely practical or necessary for a research lab, adopting specific and targeted engineering design frameworks, such as those outlined in the ISO 13485 standard, can help shift custom device development from an ad-hoc process to a more structured engineering workflow.  

Applying these industrial design controls helps establish traceability. Why were certain design decisions made? How were testing protocols developed? How is the device actually expected to function? Crucially, it forces us to systematically evaluate failure modes before a physical prototype is built, anticipating mechanical, electrical, or biological risks rather than just reacting to them on the lab bench. It moves the goal from simply getting a device to work for a specific experiment, toward designing an instrument that supports functional repeatability and outcome reproducibility.

Engineering Design Workflows for Custom Research Devices

For instance, my current project involves developing a new bioreactor system for cartilage tissue engineering. This custom device moves from an initial idea to defining clear user needs and then translating into a formal design specification document, outlining precise technical requirements and environmental constraints.  

Once those specifications are agreed upon, the detailed engineering phase kicks off, moving through component sourcing, CAD modelling, risk assessment (e.g., identifying failure modes), and prototyping. Crucially, the process includes formal check-ins: 

• Design verification is conducted to confirm that the physical build actually matches those initial engineering specifications (i.e., did we build the device right?). 
• Design validation helps ensure that the device meets its intended research use in a practical setting (i.e., did we build the right device for the job?). 

A rigorously designed device should ideally minimise operational uncertainties. But realistically, research is full of surprises. So, if contingencies occur, or if parameters need to be adjusted, having a traceable documentation trail allows future developers to troubleshoot or iterate upon the existing design much more smoothly. 

A Hybrid Approach to Research Quality and Clinical Translation

By integrating these industrial engineering workflows into biomedical research, we can maintain scientific rigour while generating the kind of traceable and quality-controlled data that commercial developers and regulatory bodies prefer to see. But we must acknowledge a fundamental friction here: the core metrics of academia and industry naturally differ. Industry replies on scalability and regulatory compliance, which require comprehensive QMS frameworks to get there, whereas research should be all about scientific discovery and knowledge generation. 

Forcing a copy-paste industrial QMS onto a university lab risks compromising the very innovation it aims to protect. It demands that researchers double as full-time quality assurance administrators, a role that drains our already limited time and funding. Standardising the engineering design workflow in a research setting is essential, but we should be careful to avoid the trap of standardising everything.  

Because of this, a strategic hybrid model could be a promising path forward. Research labs could focus limited resources on the engineering core while establishing basic technical documentation and core verification frameworks early in the proof-of-concept phase. Then, as the technology matures, research teams can collaborate strategically with accredited commercial laboratories. Managing audit-level quality and regulatory compliance becomes the focus of industrial professionals. This hybrid approach respects the realities of researchers’ constraints while leveraging industrial best practices, ultimately helping ensure that benchtop innovations are well-positioned to make a tangible impact. 

References

• J. K. Shen, T. B. Huang, C. E. Davey, et al. Standardized Effect Measures Informing Next-Generation Strategies for Mechanical Stimulation in Cartilage Tissue Engineering. Advanced Healthcare Materials 15, no. 26 (2026): e71309. https://doi.org/10.1002/adhm.71309 
• ISO 13485:2016 Medical Devices — Quality Management Systems — Requirements for Regulatory Purposes 

Disclaimer. The views and opinions expressed in this article are solely those of the author and do not necessarily reflect the official policy or position of Test Labs Limited. The content provided is for informational purposes only and is not intended to constitute legal or professional advice. Test Labs assumes no responsibility for any errors or omissions in the content of this article, nor for any actions taken in reliance thereon.