Proton Therapy as a Case Study in Systems IntegrationEngineering, Regulatory, and Operations at the Frontier
What Does Real Complexity Look Like?
The global proton therapy systems market was valued at approximately $1.49 billion in 2024 and is projected to reach $1.66 billion in 2025 and $4.09 billion by 2035, representing a compound annual growth rate of 9.4%.¹ Behind those numbers is one of the most demanding engineering and regulatory challenges in all of medical devices. A proton radiation therapy system is not a product you can hold in your hand or explain in a single sentence.
System Assembly Integration Challenge
Particle Accelerator Synchrotron physics; specialist sourcing via university and research institution
networks
100-Ton Rotating Gantry Precision mechanical engineering; facility structural integration across multiple
floors
Patient Positioning System Sub-millimeter accuracy robotics; real-time imaging feedback loop
X-Ray Imaging Assemblies Beam guidance; multi-vendor integration with treatment planning software
Treatment Planning Software Class II software as a medical device; FDA 510(k) pathway of its own
Radiation Monitoring Environmental safety; facility boundary compliance; regulatory reporting
Safety System (Overarching) Cross-assembly shutdown logic; zero-ambiguity fail-safe architecture at system
level
Purpose-Built Facility Radiation containment walls; specialized electrical, cooling, and spatial
infrastructure
All of it operating in a clinical environment where the margin for error is effectively zero.
In 2025, the American Cancer Society estimated 2.04 million new cancer diagnoses in the United States.² Proton therapy is increasingly considered for cases where conventional radiation would damage too much healthy tissue, particularly in pediatric oncology and tumors near critical structures. The demand is real, the clinical rationale is strong, and the engineering complexity that stands between concept and patient treatment is extraordinary.
Most discussions about systems integration in medical devices focus on software interfaces, regulatory checklists, or project timelines. What follows is grounded in first-hand experience bringing a proton radiation therapy system through the entire development lifecycle: FDA 510(k) clearance at a major academic medical center, and the early stages of international regulatory planning for a proton therapy facility on another continent. What made it instructive, and what makes it relevant to programs far less extreme, is that it required bridging people, process, tools, and technology simultaneously, across engineering, regulatory, and operations. The lessons are real, and many of them apply to any complex Class II or Class III device program.
The Most Consequential Decision Comes First: Defining the Framework
Before a single design decision is made, before a supplier is contacted, before a verification protocol is written, the most important work in a complex medical device program happens on a whiteboard. You have to understand what regulatory success actually means—not in the abstract, but operationally. You have to build the product structure around it from the start.
Working outside the regulatory framework reliably leads to rework. Re-design, re-verification, re-validation, re-submission: all of it costs time and money that most programs cannot absorb. The hard-won understanding from years of navigating this with a Washington, DC regulatory attorney is that the pre-market regulations are deeply interconnected. Treating them as independent checkboxes is where programs run into trouble.
For this program, approximately 20 core assemblies were defined—each consisting of hardware, software, and firmware—described with consistent nomenclature that every discipline on the program (and throughout the corporation) shared. That product structure was not just an organizational convenience. It determined how every compliance obligation downstream would be approached: design inputs, risk analysis, supplier qualification, verification, and validation.³
A critical distinction governed how those 20 assemblies were handled:
Assembly Classification Regulatory Path Documentation Approach Scope
Can operate independently Own 510(k) clearance Full independent regulatory Handful of
of all other assemblies pathway documentation; assemblies only
separate submission if needed
Cannot operate 510(k) by Deep supplier documentation Majority of
independently; supplier incorporation into review; gaps addressed assemblies
had medical-device-grade system submission through design analysis
documentation
Cannot operate Black-box approach; Internal team accepts Significant minority;
independently; supplier internal full integration requires deepest
documentation insufficient responsibility accountability; internal resource
or unavailable black-box DFMEA conducted investment
The product structure also defined boundaries of responsibility in ways that went far beyond engineering. It set bill-of-material structure, determined integration accountability, and established how sub-assemblies mapped to assemblies and ultimately to the system. Seen through the lens of people, process, tools, and technology, this foundational step was all four at once: it aligned the team around shared nomenclature, established the process logic for every downstream compliance obligation, informed the selection of appropriate tools to manage traceability, and set the technical scope each discipline was responsible for. For a system this complex, alignment on structure is not a housekeeping task. It is the foundation of everything that follows.
Engineering Integration Across 20 Assemblies
Each of the 20 assemblies—coming from all over the globe—was itself a sophisticated system, developed over years, and in many cases being used in other industries as well. The people dimension of integration was the first challenge. Finding the right level of technical expertise to understand a given assembly well enough to integrate it required lateral thinking, drawing on analogous skills from analogous industries. Finding a physicist with specific synchrotron particle accelerator experience, for example, was not a straightforward search. It required extensive networking with universities and research institutions. In practice, no amount of good process or capable tools compensates for gaps in the technical expertise needed to ask the right questions of a complex assembly supplier.
Once the internal technical team was in place, the next layer of challenges was the supplier network. Language barriers, time zones, intellectual property protections, and contractual ambiguity around integration responsibility were all real friction points. Engineers and physicists had to hold two things in mind simultaneously: a command of the most detailed technical principles of a given assembly, and a clear view of how that assembly fit into the larger integration goal.
Suppliers who had not worked in medical devices before—and that was the majority of them—were often not familiar with the design and documentation rigor that FDA 21 CFR Part 820 requires. In some cases, the internal team found itself teaching suppliers what compliance actually meant. When a supplier’s documentation was insufficient and cooperation was limited, the internal team fell back on a black-box approach for that assembly and accepted full integration responsibility.
Given the physical scale of a proton radiation therapy system—which requires special facility construction with radiation containment walls, significant electrical infrastructure, specific cooling infrastructures, and precise spatial relationships between assemblies—assembling the entire system at the factory before shipping was neither practical nor economically viable. Levels of assembly, integration, and installation were defined deliberately: what gets verified in-house, what gets verified at the destination facility, and how installation interfaces with pre-designed facility infrastructure all had to be planned in advance. The levels of verification worked together like layers of an onion, each one logically and technically supporting the next, in a way that made both engineering and regulatory sense.
Risk Management at Scale
Risk management for a system this complex is a different undertaking than for a typical device program. ISO 14971⁴ provides the framework, but the framework has to be applied across 20 major assemblies, each with its own Hazard Analysis, DFMEA, PFMEA, and UFMEA (where applicable), plus approximately 50 international safety standards mandated as Design Inputs by the FDA.
Analysis Type Scope Who Conducted It Key Output
Hazard System & All assemblies; ross-functional Identifies hazardous situations and
Analysis assembly level (engineering, physics, causal chains; foundation
clinical, service) for all downstream risk work
DFMEA Assembly level Internal team for assemblies Identifies design failure modes;
without supplier documentation; directly informs safety
supplier-provided where available system response logic
PFMEA Manufacturing & Internal engineering staff Process failure modes during
assembly assembly and installation;
supports verification planning
UFMEA Use / clinical Internal team with Use-related failure modes; directly
environment clinical input supports IEC 62366-1 usability
engineering obligations
Layered across all of this risk analysis work was a design obligation that proved to be one of the more technically demanding aspects of the entire program: the overarching system-level safety system.
Suppliers who had conducted thorough risk analyses had developed safety systems within their own assemblies—appropriate and expected. The challenge arose with assemblies that had not conducted extensive risk analysis and consequently had no embedded safety system of their own. For those assemblies, the internal team carried the full burden of safety event handling: defining what a safety event looked like at that assembly’s boundaries, and determining the appropriate system-level response—meaning what gets shut down, to what degree, and in what sequence—so that the overall system transitions to a known safe state without creating new hazards in the process.
For a system delivering radiation to cancer patients, the margin for ambiguity in that logic is zero.
Designing and verifying that architecture required a thorough understanding of every assembly’s failure modes, its response to commanded shutdown, and the timing relationships between assemblies during a safety event. It was a system-level design problem that could not be delegated to any individual assembly supplier.
The trace matrix challenge became apparent quickly. Trying to maintain traceability across 20 assemblies, each with its own set of design inputs, risk analyses, and test results, using spreadsheets is nearly unworkable at this scale. This drove the evaluation and implementation of purpose-built requirements management software with a custom architecture configured for both top-down and bottom-up traceability. When the FDA expects a clear line of evidence from user requirement to design input to risk mitigation to verification result, the tool is not a convenience. It is a structural necessity.
IEC 60601 Safety Compliance and EMI/EMC Testing: What It Actually Takes
To support the 510(k) submission,⁵ the decision was made to engage Intertek, a certified independent laboratory, to conduct both the IEC 60601⁶ safety and risk assessment and the EMI/EMC testing. Having an independent certified lab evaluate compliance against international safety standards, clause by clause, across 20 assemblies, took approximately one year. That timeline reflects the scope of the work, not inefficiency.
The EMI/EMC testing presented logistical challenges that have no equivalent in most device programs. The system was installed on an active oncology floor of a major academic medical center. EMI/EMC testing covers both radiated emissions and radiated susceptibility, and in a hospital environment, signals from the system under test can migrate to adjacent and overhead clinical areas in unintended ways. The system was surrounded by active clinical functions on adjacent floors and in rooms directly above and beside the installation.
Every individual test was mapped out in advance. The test equipment, the test location, what the test would physically do, and where signals could propagate: all of it was discussed in detail with the Intertek team before a single test was run. Each test was scheduled deliberately. This was not bureaucratic caution; it was the operational reality of conducting regulatory testing in a working hospital. The coordination required with the hospital facilities team, the clinical staff, and the test team was substantial, and the upfront planning directly enabled the testing to proceed without disruption to patient care.
The International Dimension: International Facility and TGA Regulatory Planning
The international facility project added an entirely different layer to the integration challenge. The company was positioning itself to bring the first regulatory-approved proton radiation therapy system to the Australian continent, which meant navigating the Therapeutic Goods Administration (TGA),⁷ Australia’s equivalent of the FDA, with its own regulatory structure and submission requirements.
An external Australian regulatory consultant was engaged to develop a phased plan for corporate QMS acceptance under ISO 13485 and product certification by the TGA. The gap between FDA and TGA requirements, while not insurmountable, requires specific expertise to navigate. Adding international regulatory planning to an already complex domestic program meant maintaining two parallel compliance tracks simultaneously, with remote team coordination across significant time zones.
The facility design dimension of the Australia project added another element: international facility design coordination, where the pre-designed interfaces the proton therapy system assemblies would connect to had to be specified and communicated to a construction team operating in a different country and regulatory environment. The integration requirements did not stop at the system boundary. They extended into the facility infrastructure.
What Transfers to Any Complex Device Program
The proton therapy experience is an extreme case. But the principles it forced into sharp relief apply to any serious Class II or Class III device development program:
Principle What It Means in Practice Cost of Getting It Wrong
Structure follows Build your product structure Re-design, re-verification, re-validation,
regulatory framework around the regulatory re-submission. QMSR and ISO 13485:2016
submission from day one are now co-requirements under
—not as a parallel exercise 21 CFR Part 820 (effective February 2, 2026)
Product structure is How assemblies are named, Ambiguity in product structure multiplies
a regulatory decision defined, and bounded determines risk analysis gaps, supplier accountability
how every downstream gaps, and traceability failures
compliance obligation is assigned
Supplier integration Understand what each supplier has Gaps discovered late in development become
is consistently produced, what they will share, black-box engineering problems under
underestimated and where your team must deadline pressure
accept full integration
responsibility
Risk management Purpose-built requirements Spreadsheet-based traceability across 20+
at scale requires management software with assemblies and 50+ standards is not
living tools bidirectional traceability workable at FDA inspection standards
is a structural necessity
—not a convenience
Real-environment Map every test in advance: Conducting EMI/EMC testing in an active
testing requires equipment, location, signal hospital without advance planning disrupts
pre-planned propagation, clinical adjacency patient care and invalidates results
coordination
For us, bridging people, process, tools, and technology is not a tagline. It is a description of what this program actually required, every day, across engineering, regulatory, and operations. The patterns that emerged from this experience show up in programs that are far less extreme, and they tend to respond to the same underlying approach: get the structure right early, treat risk as a living process, and understand your supplier relationships honestly before they become integration problems.
About Springboard Solutions LLC
Springboard Solutions LLC provides strategic, tactical, and regulatory consulting services for the medical device industry, specializing in systems integration, FDA 510(k) and QMSR compliance, ISO 14971 risk management, requirements traceability, and international regulatory strategy. Founded by Dan Raymond, who served simultaneously as VP of Engineering and Acting Director of Quality/Regulatory for a proton radiation therapy system, Springboard Solutions brings first-hand experience with the most complex device programs in the industry.
#SpringboardSolutions #MedicalDevices #SystemsIntegration #ProtonTherapy
References
1 Roots Analysis. Proton Therapy Systems Market: Industry Trends and Global Forecasts, till 2035. June 2025. https://www.rootsanalysis.com/reports/proton-therapy-systems-market.html
2 American Cancer Society. Cancer Facts & Figures 2025. Atlanta: American Cancer Society; 2025. https://www.cancer.org/research/cancer-facts-statistics/all-cancer-facts-figures/2025-cancer-facts-figures.html
3 U.S. Food and Drug Administration. Premarket Notification 510(k). https://www.fda.gov/medical-devices/premarket-submissions-selecting-and-preparing-correct-submission/premarket-notification-510k
4 International Organization for Standardization. ISO 14971:2019 Medical devices—Application of risk management to medical devices. Geneva: ISO; 2019. https://www.iso.org/standard/72704.html
5 Intertek Group. Medical Device Testing and Certification. https://www.intertek.com/medical/
6 International Electrotechnical Commission. IEC 60601-1:2005+AMD1:2012+AMD2:2020 Medical electrical equipment—Part 1: General requirements for basic safety and essential performance. Geneva: IEC. https://webstore.iec.ch/en/publication/67497
7 Australian Government, Therapeutic Goods Administration (TGA). Medical Devices—Overview. https://www.tga.gov.au/products/medical-devices/medical-devices-overview
8 U.S. Food and Drug Administration. Quality Management System Regulation (QMSR) Final Rule. Federal Register. February 2, 2024. Effective February 2, 2026. 21 CFR Part 820. https://www.federalregister.gov/documents/2025/12/04/2025-21955/medical-devices-quality-management-system-regulation-technical-amendments
