Physical AI: The Collision of the Deterministic OS and the Non-Deterministic LLM

Over a generation we have witnessed the maturation of the embedded market from a basic runtime environment to a safety critical layer, to a platform for AI at the edge. Let us examine "what next?" for RTOSs as we seek to maintain the reliability of the past, with the flexibility of the future.

We propose that the primary concern when choosing an RTOS are its interfaces to the rest of the system. The implementation and lack of standardisation of those interfaces have an outsize impact on the cost of using that RTOS. RTOSs have two interfaces that matter, an interface upward to application software, and one downward to the computer hardware. The upward interface must increasingly support intelligent applications that will inherently have unpredictable behaviour. The downward interface is comprised of the device driver catalogue and BSP (board support package) library that must now support not just general-purpose CPUs, but GPUs, NPUs and TPUs as the use cases for AI become more specialized.

The question we must ask is how to ensure these design parameters operate efficiently, in harmony with each other, whilst never yielding on the reliable, deterministic baseline that all mission critical systems must achieve.

History has shown us that as interface sprawl continues, market fragmentation increases leading to increased cost, complexity, lock in, and risk. Standards have demonstrated an ability to be the ultimate unifier but only if the vendors put aside personal ambition to allow the industry to align together. The enterprise world has learned this lesson to create immutable yet interoperable horizontal layers. The question is if the embedded world can achieve the same.

In this session we will discuss the pros, cons, challenges and opportunities of this thinking, and debate if the industry could, and should move there.

Speaker

Sandeep Modhvadia

Chief Product Officer, Wind River

As Wind River Chief Product Officer, Sandeep Modhvadia is responsible for driving product strategy and product management, playing a critical role in Wind River’s leadership advancing the software-defined future of mission-critical systems.

He has over two decades of technology and product management experience. Most recently at Acuity Brands, he led product management for its cloud software business as part of the Intelligent Space Group, focused on sustainability, automation, and optimization of buildings. Before Acuity, he founded and led the product and solutions team for the Google Cloud Automotive and Manufacturing group launching its platforms for connected vehicles and connected factories. Prior to Google, he held multiple product, sales, and marketing leadership positions at Microsoft.

He holds a BSc degree in computer science from University College London.