Advanced sensing, intelligent telemetry and grid intelligence for the next generation of power systems.
The problem
State estimation, high-impedance fault location, DLR and anomaly detection already rest on solid science. What is missing is measurement — in sufficient quantity, quality and frequency.
Distributed generation, renewables and electrification make power flows bidirectional and volatile. Operating with a handful of measurement points is no longer enough.
Many feeders and lines still run with limited instrumentation. Every new point costs hardware, installation, power and communication.
Streaming waveforms continuously demands bandwidth, energy and storage that field networks don't have. The bottleneck is communication, too.
“The next generation of power grids needs more than sensors. It needs a new observability infrastructure.”
Technology
Three layers designed together to reduce the total cost of observing the grid — not just the cost of a sensor.
Compact devices installed directly on the infrastructure, targeting metrological quality at competitive cost with simplified installation.
Sensor and server share a model of electrical behavior. The device mainly sends what is new — the innovation its measurement adds.
TelemetryA scalable environment ingesting conventional telemetry and high-resolution oscillography — from our sensors and, in the future, third-party devices.
Research front · Latent Electrical Telemetry
A three-phase waveform is highly predictable most of the time. We investigate neural encoders and shared dynamic models so the sensor transmits compact representations and residuals — and the server reconstructs the signal or directly extracts the information an application needs.
In steady state almost nothing is sent. When an event occurs, the residual grows and detail is preserved.
Illustrative visualization of the concept. Compression rates and performance depend on ongoing experimental validation.
Applications
Our technology roadmap. Each application depends on measurements that are scarce today — exactly the bottleneck we target.
Estimate voltages and flows with distributed, synchronized measurements, reducing reliance on pseudo-measurements.
Locate conventional and high-impedance faults from traveling waves, phasors and event signatures.
Monitor sags, harmonics, unbalance and transients along the feeder, not only at the substation.
Estimate real transmission capacity from current, conductor temperature and ambient conditions.
Track operating conditions of equipment and infrastructure for condition-based maintenance.
Identify events and atypical behavior by combining signal processing and machine learning.
Decide how many sensors, where and at what quality to reach an observability target at minimum cost.
Each application depends on measurements that are scarce today — exactly the bottleneck we target.
Observability as a project
A utility wants better fault location on a critical feeder. Rather than selling boxes, we aim to answer — with simulation and data — what it takes to get there.
Which points to instrument to make the problem observable.
Required accuracy class, sampling rate and synchronization.
Bandwidth and energy per point, with and without intelligent encoding.
Expected estimation or location error for the chosen configuration.
Deployment and operating cost against operational benefit.
Research & Development
A computational environment to validate architectures before building them — and to demonstrate their benefits quantitatively. Electromagnetic simulation, digital sensor models, communication and algorithms in a single chain.
Grid simulation
ATP/EMTP · Python
Sensor model
Noise · error · timing
Encoding
Compression · latent models
Communication
Loss · latency · bandwidth
Reconstruction
Waveforms · events
Algorithms
Estimation · faults · PQ
Assessment
Techno-economic
Planned evolution towards hardware-in-the-loop (HIL) testing.
How do sensor quantity and quality affect electrical observability?
What is the lowest communication rate that preserves a given information quality?
How can latent models exploit temporal and spatial correlations between measurements?
How do we preserve rare events under compression?
How do synchronization and communication losses affect grid algorithms?
Which sensor configuration minimizes cost for a target performance?
How do we integrate sensing, communication and intelligence in one architecture?
We develop proprietary technology and intend to publish scientific results when appropriate, respecting intellectual property.
About
We are a Brazilian technology company born at the intersection of power systems, electronic instrumentation, data communication, artificial intelligence and optimization. We don't replace protection relays from established manufacturers — we expand the quantity and quality of information available about the grid.
Transform power system observability by democratizing access to high-quality measurements through innovative, accessible and intelligent technologies.
Make it possible to observe, understand and optimize every relevant point of transmission and distribution networks.
Engineering rigor
Claims backed by physics, simulation and testing.
Hardware + software
We design the whole chain, conductor to algorithm.
Open science where possible
We publish and collaborate with academia.
Total cost, not unit price
We optimize the system, not the part.
Partnerships
We are looking for partners for R&D projects, field pilots and co-development.
Talk to us
Tell us about your grid's challenge. We'll reply with a technical conversation, not a pitch.
Send an emailcontato@g2wia.com.br
Utilities
Pilots on critical feeders and lines, regulated R&D projects.
Transmission operators
Line monitoring, DLR and fault location.
Research centers
Joint supervision, test benches, data and co-authored publications.
Manufacturers & integrators
Integrating sensors and platform into existing portfolios.
Investors
A deep-tech thesis in critical infrastructure.