Member Spotlight: Momentedge

August 7, 2026

Momentedge is building the cloud and edge intelligence layer that helps autonomous fleets learn from real-world operations, giving deployed systems a way to capture, interpret, and act on operational data at scale.

Value of Momentedge’s AUVSI Membership

Momentedge values AUVSI membership for the bridge it provides between technical development and operational reality, bringing together the operators, technology providers, policymakers, and safety leaders shaping the future of uncrewed and autonomous systems.

Melika Golkaram, Founder, Momentedge, says the company values the opportunity to learn directly from fleet operators, contribute to industry conversations about safety and reliability, and build relationships across the ecosystem as Momentedge develops intelligence infrastructure for deployed autonomous fleets.

Exciting Industry Developments

Momentedge points to the transition from individually capable machines to deployed fleets that learn continuously from real-world operations as the most exciting development on the horizon.

According to Golkaram, much of today’s physical-AI data infrastructure was designed to create training data before deployment, through teleoperation, crowdsourcing, and simulation. Operating fleets introduce a different challenge: continuously capturing, interpreting, and acting on multimodal evidence from the real world.

“Advances in multimodal AI, edge computing, and standardized data formats will enable fleets to identify patterns across video, telemetry, and sensor logs, distribute what they learn, and improve every deployed system, not only the one that experienced an event,” Golkaram says.

Industry Challenges

Momentedge identifies the gap between the volume of operational data autonomous systems generate and the speed at which teams can learn from it as the top challenge facing the industry. Golkaram notes that a critical event may last only seconds but remain buried within hours of video, telemetry, and sensor logs. Investigation is often manual, she says, and lessons remain fragmented across deployments. Scaling autonomy safely will require trustworthy ways to capture the right evidence, correlate signals, identify recurring patterns, and translate findings into action across an entire fleet.

Momentedge’s open-source Clipper preserves the context surrounding important events in standard MCAP files. The company’s commercial platform will analyze existing fleet data, identify incidents and recurring patterns, assemble supporting evidence, and recommend engineering or operational actions. Over time, validated patterns can be returned to the edge as rules or lightweight detectors, creating a feedback loop in which every incident helps improve the fleet.

Supporting High Safety Standards

Momentedge is working to make safety improvement more proactive, evidence-driven, and repeatable. The company’s open-source Clipper preserves the data surrounding incidents, anomalies, and near misses, including moments that could have become incidents, using standard, inspectable MCAP files.

Momentedge recently completed its edge recording open-source code to support ROS2 and Rust-based systems, allowing teams to choose the data streams they want included in the recorded MCAP package, fully synchronized with accurate timestamps. More on the project is available at Momentedge’s resources page.

Golkaram says Momentedge is developing cloud analysis that correlates video, telemetry, and sensor data to surface recurring patterns and weak signals across a fleet, while preserving the evidence behind each finding for human review. Because operational data can contain sensitive visual, location, and system information, the company is also working toward customer-controlled deployment options that allow data to remain within the customer’s own cloud environment. The goal, she says, is to help teams progress from reactive incident investigation to earlier risk identification and measurable fleet-wide learning.

Building Public Trust

Momentedge aims to help make continuous fleet learning a standard industry capability rather than a bespoke system available only to the largest autonomy companies. The company plans to contribute open-source edge infrastructure, support interoperable formats, and integrate with the tools robotics teams already use, reducing fragmentation across vehicles, data systems, and engineering workflows.

Through AUVSI and the broader ecosystem, Golkaram says Momentedge also hopes to contribute practical lessons around evidence capture, near-miss analysis, customer control of operational data, and responsible human oversight of AI-generated findings.

“Our role is to help fleets of every size build a repeatable practice of learning from real-world operations,” Golkaram says.