So you’ve developed a new scientific idea or principle that you believe might have some commercial potential. Perhaps you’ve taken it a step further and have a prototype on your lab bench, which – just like the first transistor  – works if you wiggle it just right. Just how far down the road is your prototype to becoming a product?
As a student, I studied particle physics, using the giant accelerator as DESY (the German equivalent of CERN) to smash electrons into protons. My project aimed to get a better understanding of the fundamental forces binding atomic particles, about as far from commercial reality as can be imagined.
After completing my D.Phil, I moved to Australia and joined a team in CSIRO that applies nuclear and X-ray physics to solve challenging measurement and imaging problems for industry. The team develops new concepts, tests them in the lab, builds prototypes that are trialled with customers, and then works with industry to turn these into commercial products.
One of the earliest projects that I worked on was a new air cargo scanning technology , combining X-ray and neutron imaging to better detect threats such as narcotics and explosives. We rapidly developed and demonstrated the physics principles needed to identify different materials in cargo, and spent the next 12 months building a prototype scanner. Sure, the image resolution was poor, and the system was too slow for real-world application, but we thought that those were straightforward scale-up issues. In 2003, we started plans for a full-scale demonstration.
Little did I know then, but we were just 18 months into a project that had another decade to run: nearly a year to plan and raise funds for the full trial; two and a half years to build and run our first system at Brisbane Airport; two years to find a commercialisation partner and then nearly five years to evolve the technology into a competitive form for international roll-out.
Along the way, almost every component of the scanner was discarded and redesigned from scratch: detector systems, radiation sources, shielding and software. Bringing in X-ray technology from our commercial partner dramatically improved image sharpness. The scanner jumped in size when a high-power neutron generator was introduced to speed up scanning, and then shrank back as a radically new detector system let us switch back to a much smaller source.
In today’s increasingly entrepreneurial climate, the language of start-up companies is becoming more common: the minimum viable product, “fail fast, fail often”, pivoting, and disruption. A common thread is that innovation needs to be fast. Build your product in a few weeks, get it in front of customers, and change or discard it if no one bites. All very well if you are building a website or a phone app, but how well does this philosophy align with science-based technologies that can take years to develop?
I’ll focus on just one of these start-up concepts: the minimal viable product or MVP, commonly defined as the smallest, simplest, cheapest and lowest-featured product that does just enough to engage customers. (In medicine, it might be a drug that just manages to cure more people than it kills – unsurprisingly, the MVP idea has never really caught on in the life sciences).
Our first full-scale cargo scanner – in hindsight, our MVP – managed to miss most of these targets. It was certainly the largest system that we ever built, and included features, such as a fancy cargo handling system, that we never used again. However, it proved to be an MVP in one important respect: with the state of technology available at the time, it was the product that we were able to get in front of potential customers in the shortest possible time. Without it, it was very unlikely that we would have generated the commercial interest, funding and even the ideas to design more effective scanning systems.
So don’t dwell on the fact that great science-based products take years to develop, but do push as hard as you can to get something in front of paying customers as early as possible. Their feedback cannot be replicated by staying locked in a lab, refining a concept for imagined users.
If you’ve come up with a working prototype, congratulations! You’re probably about 10% of the way towards turning your original idea into a product. Your scientific training has brought you so far, but the remainder of the journey is going to need new skills: talking to customers to figure out what they need (and not just what they ask for); working out whether to engineer for performance or cost and convenience; a whole new language relating to commercialisation. Much of this won’t seem like science, but is every bit as intellectually demanding, frustrating and fun. The last little 90% of road lies ahead.
 Charles Weiner (1973) “How the transistor emerged”, IEEE Spectrum 10(1) pp24-33
 CSIRO Air Cargo Scanner, https://csiropedia.csiro.au/air-cargo-scanner
About the Author:
James was born in the UK and studied physics at Oxford University. He moved to Australia in 1998 to join CSIRO, working on the development of nuclear instrumentation for the minerals and security industries. He specialises in the development of techniques for modelling radiation, and using these to invent new ways of solving challenging measurement and imaging problems. Working with industry to implement new solutions is an important part of his role.
He is a keen advocate for bringing together researchers from different disciplines, countries and particularly from industry and academia, and to this end helped found the Global Young Academy, the Australian Early-Mid Career Researchers Forum and the Australian Science and Innovation Forum.
He has received numerous awards for his work including two CSIRO medals, the Australian Academy of Science Frederick White prize, the Eureka Prize for Science in Support of Defence or National Security and an ADC Leadership Award.