Not (Yet) the Revenue Model by Susannah Dragosavac, Mewburn Ellis
The centre of gravity in surgical robotics has shifted. While early innovation focused on mechanical precision and hardware capability, the most significant advances are now increasingly software-driven. AI-enabled systems process intraoperative imaging, instrument telemetry, and workflow data in real time, enabling the interpretation of anatomy, intraoperative decision support, and continuous performance improvement. Progress is no longer measured solely in hardware refinement, but in the cognitive capabilities of the system: its ability to assist, adapt and learn.
This AI evolution might be expected to drive a corresponding shift in how value is monetised, for example leading to a more software-led revenue model (e.g. software components and subscription). In surgical robotics, however, the revenue model appears resiliently unchanged, with the dominant revenue stream being the physical products.
A clear illustration is provided by Intuitive Surgical, where over the past three years approximately 60% of revenue has been derived from instruments and accessories, around 24% from system sales, and approximately 16% from services1. The dominant contribution therefore remains the physical product (particularly the consumables used in surgical procedures), rather than the underlying software capabilities that increasingly differentiate the platform.
This apparent mismatch reflects a deeper structural feature of the sector.
Consumables provide recurring, procedure-linked revenue that scales directly with utilisation of the installed base. Once a system is adopted within a clinical setting, demand for compatible instruments and accessories becomes embedded in routine practice. This creates a predictable and durable revenue stream, reinforced by regulatory constraints and system compatibility requirements that limit third-party substitution.
From an intellectual property perspective, consumables are also particularly well suited to protection and enforcement. They are characterised by tangible, structurally defined features that map readily onto conventional patent claim formats. Infringement is typically observable and enforceable through the inspection of physical products. By contrast, AI-driven innovations, although strategically critical, can be more difficult to capture within robust and enforceable patent claims.
This gives rise to a notable paradox. AI is increasingly central to competitive differentiation, yet it does not directly underpin the dominant revenue stream. Its role is instead indirect but essential. Improvements in imaging, guidance, automation, and workflow efficiency influence system adoption decisions, expand the range of surgical indications, and enhance surgeon engagement. The economic effect is therefore mediated: AI strengthens demand for systems, which in turn drives utilisation and, ultimately, consumables revenue.
This dynamic has important IP strategy implications. Patent portfolios must support both the present and future value chain. Protection directed to instruments and accessories secures the core revenue engine. However, failure to adequately protect AI-driven functionality risks erosion of system competitiveness, with downstream consequences for the installed base and its associated recurring revenue. As a result, AI patents should be assessed not simply in isolation, but in terms of their contribution to maintaining system adoption and procedural volume.
Looking forward, there are plausible pathways through which the revenue model may evolve, even if such change has so far been limited. One avenue lies in the increasing emergence of software as a medical device (SaMD), where AI-driven modules (such as anatomical segmentation, intraoperative guidance, or predictive analytics) are approved and marketed as distinct or licensable features. This could enable recurring revenue streams more directly linked to software functionality.
Another potential shift could lie in outcome-based models. As healthcare payers move toward value-based reimbursement, and AI demonstrates clinical performance (reducing complications, shortening procedure times, or enhancing recovery), there may be increasing scope for pricing mechanisms linked to clinical value rather than procedural volume alone.
If the surgical robotics revenue model does evolve to strengthen the revenue stream from software, the strategic importance of AI patents and their enforceability may increase further.
In summary, surgical robotics currently exhibits a stable and coherent economic structure: consumables generate the majority of revenue, while AI drives differentiation and adoption. The key insight is not that the model is misaligned, but that it is layered. Hardware and consumables remain the monetisation backbone, while software acts as the strategic lever sustaining and expanding that revenue base. The future may introduce greater direct monetisation of AI, but even in its current form, AI has already become the critical control point in the competitive landscape.
1 https://isrg.intuitive.com/static-files/3606b65a-c676-47a5-a638-6b53c19b7143
