You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
feat(repo): add root capabilities and reorganize README
- add SECURITY.md with Microsoft standard vulnerability reporting
- add SUPPORT.md with community support documentation
- add dependabot.yml for pip, terraform, and github-actions
- reorganize README with frontmatter, trademark, and responsible AI
📚 - Generated by Copilot
# 🤖 Azure Robotics Reference Architecture with NVIDIA OSMO
2
-
3
-
This reference architecture provides a production-ready framework for orchestrating robotics and AI workloads on [Microsoft Azure](https://azure.microsoft.com/) using NVIDIA technologies such as [Isaac Lab](https://developer.nvidia.com/isaac/lab), [Isaac Sim](https://developer.nvidia.com/isaac/sim), and [OSMO](https://developer.nvidia.com/osmo). It demonstrates end-to-end reinforcement learning workflows, scalable training pipelines, and deployment processes with Azure-native authentication, storage, and ML services.
4
-
5
-
## 🚀 Key Features
1
+
---
2
+
title: Azure Robotics Reference Architecture with NVIDIA OSMO
3
+
description: Production-ready framework for orchestrating robotics and AI workloads on Microsoft Azure using NVIDIA Isaac Lab, Isaac Sim, and OSMO
4
+
author: Microsoft
5
+
ms.date: 2025-12-03
6
+
ms.topic: overview
7
+
keywords:
8
+
- azure
9
+
- nvidia
10
+
- osmo
11
+
- isaac lab
12
+
- isaac sim
13
+
- robotics
14
+
- reinforcement learning
15
+
- mlflow
16
+
estimated_reading_time: 5
17
+
---
18
+
19
+
A production-ready framework for orchestrating robotics and AI workloads on [Microsoft Azure](https://azure.microsoft.com/) using NVIDIA technologies such as [Isaac Lab](https://developer.nvidia.com/isaac/lab), [Isaac Sim](https://developer.nvidia.com/isaac/sim), and [OSMO](https://developer.nvidia.com/osmo). Demonstrates end-to-end reinforcement learning workflows, scalable training pipelines, and deployment processes with Azure-native authentication, storage, and ML services.
20
+
21
+
## Overview
6
22
7
23
OSMO handles workflow orchestration and job scheduling while Azure provides elastic GPU compute, persistent checkpointing, MLflow experiment tracking, and enterprise grade security.
8
24
9
-
-**Infrastructure as Code** - Terraform modules referencing [microsoft/edge-ai](https://github.com/microsoft/edge-ai) components for reproducible deployments
10
-
-**Containerized Workflows** - Docker-based Isaac Lab training with NVIDIA GPU support
11
-
-**CI/CD Integration** - Automated deployment pipelines with GitHub Actions
12
-
-**MLflow Integration** - Automatic experiment tracking and model versioning
13
-
- Automatic metric logging from SKRL agents to Azure ML
14
-
- Comprehensive tracking of episode statistics, losses, optimization metrics, and timing data
15
-
- Configurable logging intervals and metric filtering
16
-
- See [MLflow Integration Guide](docs/mlflow-integration.md) for details
17
-
-**Scalable Compute** - Auto-scaling GPU nodes based on workload demands
18
-
-**Cost Optimization** - Pay-per-use compute with automatic scaling
19
-
-**Enterprise Security** - Entra ID integration
20
-
-**Global Deployment** - Multi-region support for worldwide teams
21
-
22
-
## 🗼 Architecture Overview
23
-
24
25
This reference architecture integrates:
26
+
25
27
-**NVIDIA OSMO** - Workflow orchestration and job scheduling
26
28
-**Azure Machine Learning** - Experiment tracking and model management
27
29
-**Azure Kubernetes Service** - Software in the Loop (SIL) training
@@ -30,22 +32,27 @@ This reference architecture integrates:
-**Azure Monitor** - Comprehensive logging and metrics
32
34
33
-
**INSERT ARCHITECTURE DIAGRAM HERE**
35
+
## Quick Start
34
36
35
-
## 🌍 Real World Examples
37
+
```bash
38
+
./setup-dev.sh
39
+
```
36
40
37
-
**OSMO orchestration** on Azure enables production-scale robotics training across industries. Some examples include:
41
+
The setup script installs Python 3.11 via pyenv, creates a virtual environment at `.venv/`, and installs training dependencies.
38
42
39
-
-**Warehouse AMRs** - Train navigation policies with 1000+ parallel environments on auto-scaling AKS GPU nodes, checkpoint to Azure Storage, track experiments in Azure ML
40
-
-**Manufacturing Arms** - Develop manipulation strategies with physics-accurate simulation, leveraging Azure's global regions for distributed teams and pay-per-use GPU compute
41
-
-**Legged Robots** - Optimize locomotion policies with MLflow experiment tracking for sim-to-real transfer
42
-
-**Collaborative Robots** - Create safe interaction policies with Azure Monitor logging and metrics, enabling compliance auditing and performance diagnostics at scale
43
+
### VS Code Configuration
43
44
44
-
See [OSMO workflow examples](deploy/004-workflow/osmo/) for job configuration templates.
45
+
The workspace is configured with `python.analysis.extraPaths` pointing to `src/`, enabling imports like:
46
+
47
+
```python
48
+
from training.utils import AzureMLContext, bootstrap_azure_ml
49
+
```
50
+
51
+
Select the `.venv/bin/python` interpreter in VS Code for IntelliSense support.
45
52
46
-
##🧑🏽💻 Prerequisites and Requirements
53
+
### Prerequisites
47
54
48
-
### Required Tools
55
+
**Required Tools:**
49
56
50
57
-[pyenv](https://github.com/pyenv/pyenv)
51
58
- Python 3.11 (required by Isaac Sim 5.X)
@@ -54,72 +61,99 @@ See [OSMO workflow examples](deploy/004-workflow/osmo/) for job configuration te
| MLflow Integration | Automatic experiment tracking and model versioning ([MLflow Guide](docs/mlflow-integration.md)) |
83
+
| Scalable Compute | Auto-scaling GPU nodes based on workload demands |
84
+
| Cost Optimization | Pay-per-use compute with automatic scaling |
85
+
| Enterprise Security | Entra ID integration |
86
+
| Global Deployment | Multi-region support for worldwide teams |
87
+
88
+
## Project Structure
89
+
90
+
```text
91
+
deploy/
92
+
├── 000-prerequisites/ # Prerequisites validation and setup
93
+
├── 001-iac/ # Infrastructure as Code deployment
94
+
├── 002-setup/ # Post-infrastructure setup
95
+
├── 003-data/ # Data preparation and upload
96
+
└── 004-workflow/ # Training workflow execution
97
+
└── osmo/ # OSMO inline workflow submission
98
+
src/
99
+
└── training/ # Training code and tasks
100
+
├── scripts/ # Framework-specific training scripts
101
+
│ └── rsl_rl/ # RSL_RL training scripts
102
+
└── utils/ # Shared utilities
103
+
docs/ # Documentation and guides
70
104
```
71
105
72
-
The setup script installs Python 3.11 via pyenv, creates a virtual environment at `.venv/`, and installs training dependencies.
106
+
## Real World Examples
73
107
74
-
### VS Code Configuration
108
+
**OSMO orchestration** on Azure enables production-scale robotics training across industries:
75
109
76
-
The workspace is configured with `python.analysis.extraPaths` pointing to `src/`, enabling imports like:
110
+
-**Warehouse AMRs** - Train navigation policies with 1000+ parallel environments on auto-scaling AKS GPU nodes, checkpoint to Azure Storage, track experiments in Azure ML
111
+
-**Manufacturing Arms** - Develop manipulation strategies with physics-accurate simulation, leveraging Azure's global regions for distributed teams and pay-per-use GPU compute
112
+
-**Legged Robots** - Optimize locomotion policies with MLflow experiment tracking for sim-to-real transfer
113
+
-**Collaborative Robots** - Create safe interaction policies with Azure Monitor logging and metrics, enabling compliance auditing and performance diagnostics at scale
77
114
78
-
```python
79
-
from training.utils import AzureMLContext, bootstrap_azure_ml
80
-
```
115
+
See [OSMO workflow examples](deploy/004-workflow/osmo/) for job configuration templates.
81
116
82
-
Select the `.venv/bin/python` interpreter in VS Code for IntelliSense support
117
+
## Documentation
83
118
84
-
## 🧱 Repository Structure
119
+
| Guide | Description |
120
+
|-------|-------------|
121
+
|[MLflow Integration](docs/mlflow-integration.md)| Experiment tracking and model versioning |
122
+
|[OSMO Workflows](deploy/004-workflow/osmo/README.md)| Job submission and workflow templates |
85
123
86
-
```text
87
-
.
88
-
├── deploy/
89
-
│ ├── 000-prerequisites/ # Prerequisites validation and setup
90
-
│ ├── 001-iac/ # Infrastructure as Code deployment
│ └── osmo/ # OSMO inline workflow submission (see osmo/README.md)
96
-
├── src/
97
-
│ ├── terraform/ # Infrastructure as Code
98
-
│ │ └── modules/ # Reusable Terraform modules
99
-
│ └── training/ # Training code and tasks
100
-
│ ├── common/ # Shared utilities
101
-
│ ├── scripts/ # Framework-specific training scripts configured for Azure services
102
-
│ │ ├── rsl_rl/ # RSL_RL training scripts
103
-
│ │ ├── skrl/ # SKRL training scripts
104
-
│ └── tasks/ # Placeholder for Isaac Lab training tasks
105
-
```
124
+
## Contributing
106
125
107
-
## 🪪 License
126
+
We appreciate contributions! Check out [open issues](https://github.com/Azure-Samples/azure-nvidia-robotics-reference-architecture/issues) to get started.
108
127
109
-
This project is licensed under the MIT License. See [LICENSE.md](LICENSE.md) for details.
128
+
## Acknowledgments
110
129
111
-
## 🤝 Support
130
+
This reference architecture builds upon:
112
131
113
-
For issues and questions:
132
+
-[microsoft/edge-ai](https://github.com/microsoft/edge-ai) - Edge AI infrastructure components
133
+
-[NVIDIA Isaac Lab](https://github.com/isaac-sim/IsaacLab) - RL task framework
134
+
-[NVIDIA Isaac Sim](https://developer.nvidia.com/isaac-sim) - Physics simulation
Microsoft encourages customers to review its Responsible AI Standard when developing AI-enabled systems to ensure ethical, safe, and inclusive AI practices. Learn more at [Microsoft's Responsible AI](https://www.microsoft.com/ai/responsible-ai).
118
140
119
-
This reference architecture builds upon:
141
+
## Legal
142
+
143
+
This project is licensed under the [MIT License](./LICENSE.md).
144
+
145
+
**Security:** See [SECURITY.md](./SECURITY.md) for security policy and reporting vulnerabilities.
146
+
147
+
## Trademark Notice
148
+
149
+
> This project may contain trademarks or logos for projects, products, or services. Authorized use of Microsoft
150
+
> trademarks or logos is subject to and must follow Microsoft's Trademark & Brand Guidelines. Use of Microsoft trademarks or logos in
151
+
> modified versions of this project must not cause confusion or imply Microsoft sponsorship. Any use of third-party trademarks or
152
+
> logos are subject to those third-party's policies.
153
+
154
+
---
120
155
121
-
*[microsoft/edge-ai](https://github.com/microsoft/edge-ai) - Edge AI infrastructure components
122
-
*[NVIDIA Isaac Lab](https://github.com/isaac-sim/IsaacLab) - RL task framework
123
-
*[NVIDIA Isaac Sim](https://developer.nvidia.com/isaac-sim) - Physics simulation
Microsoft takes the security of our software products and services seriously, which includes all source code repositories managed through our GitHub organizations, which include [Microsoft](https://github.com/Microsoft), [Azure](https://github.com/Azure), [DotNet](https://github.com/dotnet), [AspNet](https://github.com/aspnet) and [Xamarin](https://github.com/xamarin).
6
+
7
+
If you believe you have found a security vulnerability in any Microsoft-owned repository that meets [Microsoft's definition of a security vulnerability](https://aka.ms/security.md/definition), please report it to us as described below.
8
+
9
+
## Reporting Security Issues
10
+
11
+
**Please do not report security vulnerabilities through public GitHub issues.**
12
+
13
+
Instead, please report them to the Microsoft Security Response Center (MSRC) at [https://msrc.microsoft.com/create-report](https://aka.ms/security.md/msrc/create-report).
14
+
15
+
If you prefer to submit without logging in, send email to [[email protected]](mailto:[email protected]). If possible, encrypt your message with our PGP key; please download it from the [Microsoft Security Response Center PGP Key page](https://aka.ms/security.md/msrc/pgp).
16
+
17
+
You should receive a response within 24 hours. If for some reason you do not, please follow up via email to ensure we received your original message. Additional information can be found at [microsoft.com/msrc](https://www.microsoft.com/msrc).
18
+
19
+
Please include the requested information listed below (as much as you can provide) to help us better understand the nature and scope of the possible issue:
20
+
21
+
- Type of issue (e.g. buffer overflow, SQL injection, cross-site scripting, etc.)
22
+
- Full paths of source file(s) related to the manifestation of the issue
23
+
- The location of the affected source code (tag/branch/commit or direct URL)
24
+
- Any special configuration required to reproduce the issue
25
+
- Step-by-step instructions to reproduce the issue
26
+
- Proof-of-concept or exploit code (if possible)
27
+
- Impact of the issue, including how an attacker might exploit the issue
28
+
29
+
This information will help us triage your report more quickly.
30
+
31
+
If you are reporting for a bug bounty, more complete reports can contribute to a higher bounty award. Please visit our [Microsoft Bug Bounty Program](https://aka.ms/security.md/msrc/bounty) page for more details about our active programs.
32
+
33
+
## Preferred Languages
34
+
35
+
We prefer all communications to be in English.
36
+
37
+
## Policy
38
+
39
+
Microsoft follows the principle of [Coordinated Vulnerability Disclosure](https://aka.ms/security.md/cvd).
0 commit comments