Kawasaki builds scalable robotics as a service platform with Azure Kubernetes Service

Building a platform for the next era of robotics

For decades, industrial robotics has been defined by specialized engineering. Robots were designed and deployed on a project-by-project basis. The model delivered powerful automation capabilities but made it difficult to scale innovation across industries and use cases. At the same time, labor shortages and aging populations were increasing demand for healthcare and elder-care support. Organizations were looking for new ways to automate repetitive work while enabling people to focus on higher-value tasks.

Kawasaki Heavy Industries recognized that robotics would need to move beyond traditional manufacturing environments and operate alongside people. The challenge was not simply building more capable robots. It was creating a platform to support intelligent robotics applications across more environments and use cases. Kawasaki set out to develop a cloud-based robotics as a service platform capable of supporting repeatable, production-scale deployment of intelligent robotics systems. The objective was clear: transform robotics from a series of engineering projects into a scalable business capability. 

Creating a modern foundation for intelligent robotics

Supporting software-defined robotics at scale required a unified architecture capable of connecting AI, operational data, and applications across the full robotics lifecycle. To achieve this, Kawasaki built its platform on Microsoft Azure, incorporating: 

• Azure Kubernetes Service (AKS) 

• Azure OpenAI in Foundry Models

• Azure Cosmos DB

• Azure IoT Operations

• Azure Digital Twins

Together, these services provide the foundation for developing, deploying, operating, and continuously improving intelligent robotics applications. Jun Yamaguchi, Director, Kawasaki Physical AI Center at Kawasaki Heavy Industries, explains, “Azure Kubernetes Service gave us a consistent platform for deploying robotics applications across environments. Instead of rebuilding software for each implementation, we can develop capabilities once and scale them across deployments.”

This approach allows AI skills to be developed once and reused across robots, enabling new capabilities to be rolled out quickly without rebuilding each application. AKS provides a consistent runtime for AI inference, application logic, and device integration, so robotics teams can deploy and update workloads across cloud and edge environments without managing different orchestration stacks.

Hiroaki Kagaya, General Manager of the Presidential Project Management Division, and Senior Manager of the Social Robot Business Strategy Department, at Kawasaki Heavy Industries, says, “Through our co-engineering collaboration with Microsoft’s Forward Deployed Engineering team, we built the platform foundation needed to bring intelligent, AI-enabled robotics into healthcare environments while creating a model that can scale to additional industries over time.”

A key part of this approach is making robotics accessible beyond specialized engineering teams. With Kawasaki’s KARAKURI initiative, users without robotics expertise—including healthcare professionals—can build and operate robot applications using natural language instructions. 

From projects to platforms

One of the most significant shifts in Kawasaki’s strategy was separating software from individual robot deployments. Historically, robotics solutions often require substantial customization for each environment. Knowledge, integrations, and workflows were tightly coupled with individual projects. Kawasaki’s Azure-based architecture enables a different model. 

Applications are packaged as reusable containerized services running on AKS. Operational data, telemetry, and application state are managed through Azure Cosmos DB, providing a centralized, continuously updated data layer that supports consistent operation and improvement across deployments. Azure OpenAI provides natural language interaction capabilities, while Azure Digital Twins create digital representations of physical environments and systems.

Explains Fumihiro Honda, Chief Executive Staff Officer, DX Strategy Division at Kawasaki Heavy Industries, “Our platform brings together AI models, application logic, data, and infrastructure in a common cloud-native architecture. Azure technologies, including Azure Kubernetes Service, provide key building blocks that support deployment and operation across multiple environments.”

This architecture allows capabilities to be developed once and reused across deployments, reducing complexity and accelerating innovation while supporting continuous improvement. 

Bringing intelligent robotics into human-centered environments

Kawasaki’s platform strategy is closely tied to its vision for healthcare, service, and social robotics. Healthcare environments are dynamic and highly variable. Systems must be reliable, adaptable, and easy to operate while supporting people rather than replacing them. 

Operational telemetry, software updates, AI services, and troubleshooting workflows can be managed through a common platform. Knowledge and operational data can be shared across deployments, supporting continuous improvement. Yamaguchi notes, “In healthcare spaces, small efficiencies really add up. With Azure OpenAI in Foundry Models, we can make operational knowledge available right when it’s needed, which helps reduce the day-to-day burden on staff.” 

In one case, where robots reduce routine workload and allow staff to focus on patient care, service robots have cut the distance nurses walk from 20 kilometers a day to 10 kilometers. 

Yamaguchi explains, “Healthcare environments require systems that can adapt to changing conditions while remaining reliable and easy to manage.” These deployments represent a broader evolution in robotics, laying the groundwork for systems that increasingly resemble what is now described as physical AI. 

From platform strategy to operating model

The shift to an operating model became visible when Kawasaki publicly introduced its Digital Platform strategy as part of its broader DX transformation and Group Vision 2030. The platform connects robotics applications, data and AI services, digital twins, and ecosystem participants through a common cloud foundation. Rather than a single product launch, it represents the ongoing development of a repeatable framework for intelligent robotics systems at scale.

By positioning the platform at the center of its future robotics strategy, Kawasaki shifted from individual deployments to an operating model where applications, data, and operations evolve together.

Governance, visibility, and continuous improvement

As robotics systems expand across environments, governance and visibility become increasingly important. Kawasaki’s platform centralizes operational data, AI-powered services, telemetry, and application management within Azure. This enables teams to monitor deployments and manage applications consistently across environments.

Azure Digital Twins further supports operational visibility by creating digital representations of physical systems for simulation, monitoring, predictive maintenance, and troubleshooting. Digital twin and simulation technologies also support application validation and system evaluation before deployment, helping bridge the gap between AI development and real-world operations. Together, these capabilities provide a replicable mechanism for deploying intelligent robotics systems consistently across environments.

A foundation for robotics as a service

Kawasaki’s transformation reflects a broader shift across manufacturing and intelligent systems. As products become increasingly software-defined and AI-enabled, competitive advantage depends less on individual applications and more on the quality of the platform supporting them. 

By moving from project-based robotics delivery toward a cloud-based robotics as a service model, Kawasaki is currently developing a foundation intended to support future growth across industrial, healthcare, service, and social robotics markets. Says Kagaya, “Our robotics as a service strategy depends on a platform that can connect data, AI, applications, and operations. Microsoft Azure provides the foundation that allows us to scale that model across industries and environments.”

For organizations exploring how to scale intelligent systems, Kawasaki's experience illustrates a broader lesson: AI alone does not create transformation. Sustainable innovation requires a modern foundation that connects intelligence, applications, operations, and data into a system that continuously learns and improves. The company continues to invest in technologies that support the evolution toward physical AI, cloud-connected robotics, digital twins, and intelligent automation as part of its long-term vision. This investment includes the establishment of the new Kawasaki Physical AI Center in San Jose, California, which aims to accelerate Japan–US collaboration in the AI and semiconductor fields. 

Kawasaki demonstrates how a unified platform approach can help transform robotics from a bespoke engineering effort into a scalable business capability.

Kagaya concludes, “At Kawasaki Heavy Industries, we are advancing the realization of physical AI by connecting robotics, trusted models, simulation, data, and control systems with Microsoft on a unified platform. The healthcare work is an important proof point: a foundation that can scale from connected robots into broader physical-world transformation.”

Discover more about Kawasaki Heavy Industries on LinkedIn and YouTube.