By Robbie Elliott

Introduction: The Overlooked Half of the Product Lifecycle

Manufacturers of complex medical devices direct considerable energy toward the moment of the sale. The engineering breakthroughs, regulatory clearances, and commercial effort required to place a new system in a hospital represent a significant undertaking. However, the sale is not the finish line of the product journey; in many ways, it is the start. The years that follow – during which the devices must be maintained, calibrated, and repaired – are where uptime, customer trust, and a substantial share of lifetime revenue are ultimately won or lost. Amid the excitement of product innovation, the network that keeps these devices running often goes overlooked.

The networks that support service and repair activities are often an afterthought – accumulating gradually rather than by design. Adding a depot to remedy long turnaround times, hiring an extra field service engineer to reduce a backlog, or absorbing new repair capability through an acquisition can appear sufficient in the short term, until an organization finds itself operating an accidental network rather than one built intentionally. The result is fragmentation: inconsistent repair quality, duplicated capabilities, uneven turnaround times, and a high cost base.

The consequences of this inattention are growing as devices are becoming more complex and customers are justifiably less tolerant of downtime. In a hospital, for example, an idle system is not merely an inconvenience but a significant cost driver. Operating-room time has been estimated to cost hospitals on average $35–40 per minute, with some analyses placing it considerably higher.^1,2 In their ongoing drive to reach more patients and improve patient outcomes, forward-thinking manufacturers are recognizing that how they structure their service network is not a back-office concern but a strategic lever – one that directly shapes quality, compliance, cost, and the customer relationship. The most effective answer is rarely a single delivery model. Rather, it is a deliberately designed network that balances the benefits of scale against the local support that customers depend on.

The Three Models

Before a network can be designed, its building blocks must be understood. Nearly every service and repair network is assembled from three underlying models: Field Service Engineers, Localized Depots, and Centers of Excellence. Each model carries a distinct profile of cost, quality, speed, and scalability.

1. Field Service Engineers (FSEs)

FSEs perform repairs on-site, at the customer’s location. For installed or non-shippable equipment – e.g., a surgical robot or hospital bed that cannot be easily boxed up and sent away – this is frequently the only practical option. Customers instinctively favor this model, as it spares them the disruption of surrendering their equipment, but it is also the most expensive and least scalable of the three. Each dispatch is commonly estimated to cost hundreds of dollars for the repair alone and can approach $1,000+ once travel, labor, and overhead are fully accounted for.³ Those costs multiply whenever a technician arrives without the correct part or expertise and a second visit becomes necessary. Additionally, since talent is distributed thinly across a wide geography, maintaining consistent quality across the workforce is a persistent challenge.

2. Localized Depots

Localized depots shift repair off-site into regional facilities to which customers return their equipment. By drawing work into a controlled environment, depots improve repair quality and reduce cost relative to field service, while remaining close enough to the installed base to keep turnaround quick. Their limitation is dilution: a dense network of smaller depots fragments repair volume across many sites, duplicating tooling and inventory and leaving specialists too thinly spread to build deep expertise in any single location or repair type.

3. Centers of Excellence (CoEs)

CoEs consolidate service and repair activities into a smaller number of larger regional or global hubs. By concentrating volume, they achieve the strongest economies of scale, the deepest technical expertise, and the most consistent, tightly controlled quality. These advantages prove decisive for complex, highly regulated devices. What they concede is proximity: a centralized hub sits farther, on average, from any given customer, complicating the logistics required to move equipment to and from the repair location.

Examined side by side (Figure 1), the trade-offs come into focus. FSEs lead on proximity but trail on cost and scale; depots occupy the middle ground; CoEs lead on cost, quality, and scale but must actively manage the distance they introduce between the repair and the customer. No single model prevails across every dimension, which is why the most effective networks combine them.

Figure 1. Comparing Field Service Engineers, Localized Depots, and Centers of Excellence

The Hub-and-Spoke Blueprint

Combining all three models into a tiered system is the blueprint of the hub-and-spoke network: a deliberately layered structure that matches each repair to the point best suited to perform it (Figure 2). The tiers form an escalation path. At the outer edge sit remote diagnostics and field service engineers, resolving or triaging issues where the equipment lives and handling the installed, non-shippable, and time-critical work that cannot wait. Routine, shippable repairs flow inward to regional depots (the “spokes”), which manage volume efficiently and close to the customer. The most demanding work – board-level repair, precision calibration, validation, refurbishment, and the servicing of newly launched products – escalates to the Centers of Excellence (the “hubs”), where specialized talent and tooling are concentrated. As used here, the model is defined by capability rather than by the central sorting facility common in logistics-heavy hub-and-spoke models.

Figure 2. Hub-and-Spoke Structure

Critically, the value of the model lies not in its tiers but in their proportions. The ratio of hubs to spokes is itself a strategic decision that should flex with a company’s size, the density of its installed base, the complexity of its products, its customers’ tolerance for downtime, and the regulatory intensity of its markets. A manufacturer of high-acuity surgical robots serving a concentrated set of academic medical centers will benefit from a different network than one supporting a broad, dispersed fleet of general instruments.

In practice, company size exerts a particularly strong influence over model design. Smaller manufacturers, with thinner installed bases, often find consolidation not merely preferable but necessary. With limited repair volume, a single Center of Excellence may be the only way to reach the scale at which specialized capability becomes viable. The underlying economics are well established – the expected cost savings from pooling demand across fewer locations rise with the square root of the number of sites consolidated.⁴ Larger manufacturers, by contrast, generate enough volume to justify a fuller, multi-tier network, and with it the freedom to tune the hub-to-spoke ratio to their strategy.

This design is not merely theoretical. Leading device makers already operate blended networks of this kind. Stryker’s ProCare program pairs on-site field service with centralized depot repair and a pool of loaner equipment that keeps customers running while their devices are serviced.⁵ Intuitive, maker of the da Vinci surgical system, combines a worldwide field-service force with regional technical support and an advance-exchange program that ships replacement components overnight.⁶ In each case, the architecture is the product of deliberate choices about what to centralize, what to distribute, and where to meet the customer.

The model also varies in sophistication. In its foundational form, the hub concentrates capability and enforces common standards while work escalates tier by tier as complexity requires. In its most advanced form, the hub becomes an active control point – enabled by a mature ERP and repair-workflow system – that registers every incoming request and dynamically routes each repair by cost, capacity, lead time, and logistics. Most manufacturers begin with the former and layer on the latter as volume, data, and systems mature.

Benefits of a Well-Designed Network

A well-designed network delivers measurable operational benefits across four primary dimensions: repair quality, equipment uptime, regulatory compliance, and new-product readiness.

Repair quality and consistency. Concentrating volume builds expertise. A technician who repairs the same subsystem hundreds of times, in a controlled environment with full tooling and parts on hand, resolves problems more reliably than a generalist. Mastery shows up as fewer repeat repairs and tighter, more predictable quality.

Equipment uptime. Uptime is the outcome customers feel most directly. Remote diagnostics resolve or narrow many issues before a dispatch is required, and when physical repair is unavoidable, loaner and advance-exchange programs keep the customer operating while the work is done. Given how quickly an idle system translates into cancelled appointments and lost revenue, protecting uptime is often the most valued outcome a service organization can deliver to its customers.

Regulatory compliance. For medical devices, repair is a regulated activity. Since February 2026, the FDA’s Quality Management System Regulation (QMSR) has aligned U.S. requirements with the international ISO 13485 standard by incorporating it directly into federal regulation.⁷ Meeting the resulting obligations – standardized work instructions, calibration control, traceability, and audit-ready records – is far more achievable when repair is concentrated in a few validated centers than when it must be replicated across dozens of dispersed sites.

New-product readiness. When a new platform launches, service capability must be ready almost immediately, yet the expertise to repair it is initially scarce. Training a concentrated team and equipping a single site is faster and more consistent than diffusing that capability across a large field force at once. The Center of Excellence becomes the launch point for service readiness, cascading capability outward as the installed base grows.

While these benefits are not guaranteed, most problems that arise are addressable when anticipated and managed deliberately. The most common objection to consolidation is the downtime introduced when equipment must travel to a distant hub, but a well-designed network can remedy this using loaner pools and advance-exchange programs. Other common hurdles tend to be organizational rather than technical – local resistance, change fatigue, and inconsistent metrics – which can be overcome through thoughtful leadership.

Conclusion: The Best of All Three

The manufacturers that will win the next decade of service are those that stop treating their networks as a byproduct of history and start engineering them as a source of advantage. The hub-and-spoke model serves as a successful design because it reconciles competing priorities. Rather than trade cost for quality or proximity for scale, it captures each where it matters most – keeping repair close and responsive through depots and field service, while concentrating deep expertise, consistency, and compliance in Centers of Excellence. The result is a network that is simultaneously efficient and resilient, standardized and local, built to scale as the installed base grows, and adaptable as the product mix evolves. For the organizations that get it right, service ceases to be a cost of doing business and becomes a driver of loyalty, growth, and competitive strength.

About the Author:

Robbie Elliott   |   relliott@accelmg.com

Robbie Elliott is a Manager with Accel Management Group. Mr. Elliott is an ex-McKinsey consultant with nearly 5 years of experience spanning Operations Strategy, Supply Chain Management, and Innovation Capability Building across Healthcare, Automotive, Government, and Private Equity applications. He graduated Summa Cum Laude from the University of Michigan with BSE and MSE degrees in Industrial and Operations Engineering. Connect with him on LinkedIn at https://www.linkedin.com/in/robbieaelliott/

Sources:

[1] Childers CP, Maggard-Gibbons M. “Understanding Costs of Care in the Operating Room.” JAMA Surgery, 2018

[2] Smith TS, Evans J, Moriel K, Tihista M, Bacak C, Dunn J, Rajani R, Childs B. “The Cost of OR Time is $46.04 per Minute.” Journal of Orthopaedic Business, 2022

[3] Technology & Services Industry Association (TSIA), field-service cost benchmarks (industry estimate, widely cited)

[4] Eppen GD. “Effects of Centralization on Expected Costs in a Multi-Location Newsboy Problem.” Management Science, 1979

[5] Stryker Corporation. ProCare service program (company materials, stryker.com).

[6] Intuitive Surgical, Inc. da Vinci services and support (company materials, intuitive.com)

[7] U.S. Food & Drug Administration. “Medical Devices; Quality System Regulation Amendments,” Final Rule. 89 FR 7523 (Feb. 2, 2024); effective Feb. 2, 2026