Spranic builds spatial memory layer for physical AI
Munich-based Spranic is developing a spatial memory layer that records physical spaces and their changes over time. The startup aims to help robots interpret and adapt to dynamic environments by connecting this data with knowledge graphs, according to founder and CEO Jordi Spranger.
The approach targets a core challenge in physical AI: enabling systems to move beyond rigid, pre-programmed tasks. While most AI safety debates focus on software alignment, Spranic’s work highlights the risks and opportunities of AI operating in physical spaces—a theme StartupReader has explored in recent coverage of embedded AI chips and Asian robotics startups.
Spranic’s framework suggests a shift toward context-aware robotics, though its commercial viability remains unproven. The startup’s focus on spatial memory aligns with broader industry efforts to improve robots’ adaptability, but whether knowledge graphs can scale for real-world use cases is an open question. For now, the company operates in stealth, with no disclosed funding or product timeline.
Sources: siliconangle.com
“Spranic’s spatial memory layer could help robots adapt to changing environments, addressing a key gap in physical AI.”
Read the original reporting
The outlets below did the original reporting.
- Spranic builds spatial memory for physical AI — siliconangle.com
Related briefs
This brief was drafted automatically from the sources above and published under our editorial policy. Spotted an error? Tell us.