SPP 0.1.0 Experimental Preview
Pre-releaseSPP 0.1.0 Experimental Preview
Purpose
This release candidate presents the Spatial Policy Protocol (SPP) as an experimental open interoperability layer between autonomous systems and physical environments. A place publishes machine-readable requirements, a machine demonstrates conformance, and the resulting evidence supports a spatially scoped operating profile.
SPP is developed under the PlaceAuth project. The robot application remains unchanged while the place requirements determine which actions are permitted, conditional, or denied.
Current architecture
- SPP Core — place requirements, hierarchical spaces, inherited policy, action families, and
permit/deny/conditionaldecisions. - SPP Conformance (experimental) — deterministic conformance plans, requirement-to-proof mapping, evidence, and assurance levels.
- SPP Admission (experimental) — evidence verification,
ADMITTED/DEGRADED/DENIEDprofiles,RequirementDelta, and selective requalification.
Implemented capabilities
- JSON Schemas and YAML examples for common physical environments.
- Local Python and OPA/Rego policy evaluation.
- Hierarchical policy inheritance with deny-by-default behavior.
- Deterministic evidence generation and binding.
- Evidence digest, freshness, policy/environment binding, and reference replay checks.
- Explicit degraded restrictions and essential-safety denial.
- ROS 2 enforcement-point stub and Docker Compose reference stack.
Four admission scenarios
python demo/admission/run_demo.py a # full conformance -> ADMITTED
python demo/admission/run_demo.py b # video-retention failure -> DEGRADED
python demo/admission/run_demo.py c # essential safety failure -> DENIED
python demo/admission/run_demo.py d # transition, reuse, and selective requalificationScenario D reuses the lobby movement guarantee and runs only the three unresolved patient-wing checks before issuing ADMITTED.
Test status
The reference environment reports 35 passed, 1 skipped, 0 failed with the full pytest suite.
Limitations and security disclaimer
This is an experimental preview, not an adopted standard or production security system. SPP does not itself force a malicious autonomous system to obey an operating profile. Trust depends on evidence assurance, enforcement-point integrity, robot/runtime integrity, site infrastructure, hardware guarantees, and any attestation or independent observation mechanisms used by a deployment.
Not production-ready: hardware attestation, distributed replay protection, production ROS 2/Nav2 or Open-RMF integration, certification infrastructure, discovery, production identity/PKI, physical enforcement guarantees, and broad vendor interoperability testing.
Contributing
See CONTRIBUTING.md, SECURITY.md, and ROADMAP.md. Reproduce the demos, review the schemas, report implementation bugs, and propose interoperability improvements with focused tests where possible.
Licensed under Apache-2.0. Certain technologies described in this project are patent pending.