Measurement. Data. Control.

From measurement to intelligent control.

We connect measurements, operational systems, GIS, time series, hardware, analytics and AI into one functional whole. We create solutions around the real energy environment – from data unification and validation through balance calculations and monitoring to prediction, recommendations and safely introduced automation.

Physical network. One data model.

Collecting data is not enough. It has to make sense.

Energy operations are rarely created as a single system. Measurements, control technology, maps, commercial data and operational records come from different environments, at different intervals and with varying quality. On their own they show values; only when connected do they create a picture of the whole.

We therefore build a coherent data layer and a model of the real physical infrastructure above the existing sources. Every value has a location, meaning, history and relationship to surrounding points. This allows data to be compared correctly, assembled into balances, displayed in GIS and used to trace a deviation to where it actually originates.

We give the same attention to the quality of the measurement itself. Communication Watchdog, physical-limit checks and long-term sensor behaviour help distinguish a change in the network from a data error. Analytics, alerts and further automation become trustworthy only when built on this foundation.

Integration and time series

One coherent data layer

We connect SCADA, control systems, BMS, GIS, remote metering databases, IoT, commercial systems and standalone meters from different manufacturers and generations. The new platform does not have to replace everything the customer already uses. We create a shared layer above the existing technology that unifies the meaning, history and frequency of data as well as the responsibility of individual sources. It can work with continuous measurements, different intervals, delayed values and long histories while preserving the correct time context.

Model and GIS

A digital representation of the real network

We always place a value in its physical and spatial context. The system knows where a measurement originated, what it represents, which branch or location it belongs to and how it relates to surrounding points. Network, area, branch, section, station, device and sensor form one logical hierarchy in which inputs, outputs and relationships can be followed. GIS is therefore not merely a map, but part of the data model that helps display network condition, locate a deviation and assemble operational calculations correctly.

Data Quality and Watchdog

Trust begins with measurement

We monitor not only the value itself, but also whether the measurement can be trusted. Communication failures, missing intervals, delayed or stuck data, unrealistic jumps, unusual noise and long-term sensor drift can all be detected automatically. Physical validation rules also verify whether a combination of pressure, temperature, flow and other quantities makes sense in the current mode. The check can consider the point’s own history, the state of nearby devices, outdoor temperature, season and technology type, making it easier to distinguish a measurement fault from a real operational change.

Balances and analytics

Calculations grounded in real operations

We combine individual measurements into logical nodes, branches and higher-level units. We compare supplied and consumed quantities, follow differences over time and help locate the part of the system where the real state begins to diverge from what is expected. Calculations can use different time windows and combine measured and derived quantities according to the physical model. Time series, relationships and network topology then form a shared basis for balances, losses and further operational analysis.

Supervision and workflow

An alert that leads to action

An alert does not end with a red icon. We classify the event, determine its priority and pass it to the right role together with context, history and input for the next step. An operational workflow can combine device condition, a physical deviation, a balance difference and an unusual trend. It also records acceptance of the event, the way it was handled and the outcome, preserving traceable responsibility and experience for later evaluation. The aim is not to overwhelm users, but to focus their attention on situations that genuinely require it.

Context. Prediction. Safe action.

AI grounded in real operations, not isolated data.

We build AI on history, time series, the physical model and the current state of surrounding devices. In this context, it can help identify an anomaly, explain possible causes, determine priority and prepare the next step. The path leads from better supervision through prediction and recommendations to the automation of selected processes.

01: Built on trusted dataAI is not a substitute for data quality or a physical model. It works with verified time series, clear relationships between devices and the rules of a specific energy environment. It therefore does not start from an isolated value, but from data whose origin, meaning, quality and relationship to the real physical process are known.
02: Finds relationships across time and networkIt compares the current condition with history, surrounding points, operating mode, weather and other inputs. It can continuously monitor large numbers of points at once and flag a gradual change or combination of parameters that could easily disappear when viewed separately. It can also explain the same situation differently for dispatch, technical operations or management.
03: Predicts and recommendsAcross long-term development it can look for risk conditions, support predictive maintenance and prepare recommendations for operators and management. It may follow gradual deterioration of parameters, changes in pressure or flow behaviour, sensor drift or correlations between several signals. This helps target inspection and maintenance according to real condition and risk, while the output remains an explainable basis for an expert decision.
04: Control is introduced in verified stepsWe add automation where data quality is sufficient, limits are known and responsibilities are defined unambiguously. Every step must have clear safety conditions, a means of control and a defined way to hand the situation over to a person. Recommendations can gradually progress to semi-autonomous control of selected processes, always respecting physical constraints and the safety of real operations.

Hardware. Evolution. Long-term operations.

From measurement to control. Step by step.

An energy system does not end with a database or dashboard. Depending on the environment, it may include proprietary or integrated hardware, sensor communication, data processing at the edge and the safe transfer of a decision back into the physical process. We therefore design software and hardware in direct relation to each other.

The scope of the solution is determined not by a list of technologies, but by the customer’s real situation and goal. In one environment, the first need may be to unify sources and verify measurement quality; in another, the next step can already involve balances, prediction or automation. We add capabilities so that every stage delivers value on its own while preparing space for the next step.

We design the entire architecture for long-term operation. We anticipate new devices, growing volumes of time-series data, changes in surrounding technology and the continued development of analytics and AI. The result is not a one-off tool, but a system that can be maintained safely, extended and gradually moved towards a higher level of control.

Physical layer

Software and hardware as one system

We can connect our own software development with electronics design, firmware, communication units, sensors and edge logic. Data can be processed at the source, events handled in real time, information transferred securely to the central model and a verified decision returned to the real environment. Our experience with proprietary electronics and advanced communication protocols enables us to design the physical and data layers together. Where needed, we coordinate specialist installation, electrical work and measurement technology with expert partners.

A meaningful start

The first step follows the real need

Every operation starts from a different point and does not necessarily need AI or automatic control immediately. The greatest first benefit may be visibility of available sources, consistent data meaning or confidence that measurements reflect reality. We therefore begin by mapping technology, data flows, physical relationships, responsibilities and the questions the system genuinely needs to answer. Together, we identify a step that creates practical value in the first stage and forms a strong foundation for further development.

Gradual development

From visibility to safe action

We develop the system in verified layers. First we see and unify the data; then we understand it in the context of the physical network, draw attention to important deviations, predict future development and prepare recommendations. Every stage has its own verifiable benefit and provides experience for designing the next layer. Automation and control of selected processes are added only where data is reliable, limits are known and responsibility is clear.

Long-term responsibility

The system must work after the project ends

Security, monitoring, updates and scalability are designed into the system from the beginning. After deployment, we monitor communication, availability, performance and data quality, as well as the condition of the measurement and communication infrastructure itself. Based on actual use, we add new measurement points, sources, calculations and integrations without losing traceability of changes. Operational experience becomes a basis for further analytics, AI and automation rather than the project ending as a one-off delivery.

Experience. Teplárny Brno.

Large-scale energy data in real-world operation.

For Teplárny Brno, we created a system that unified operational data, modelled the hot-water network and evaluated heat losses online. The solution connected data from several independent sources, worked with more than one thousand sensors and linked time series to a GIS model of the real distribution network.

The historical implementation processed flows, pressures, temperatures and supplied heat, using regular calculations to evaluate balances and losses across the real network topology. GIS visualisation, interactive analytics, data-availability supervision, sensor-fault monitoring, alerts and summary outputs gave technical roles a shared view of an extensive physical system.

This experience enabled us to verify over the long term what a large-scale energy system needs in daily operations. We are now building on it with a new generation of the central data and monitoring platform. It brings old and new measurements into a shared data model with history, Data Quality, communication Watchdog, physical validation rules, alerts and a technical monitoring console. This creates a robust foundation for further balance, predictive and AI capabilities.

1,100+: sensors in the implemented system
760: heat-exchange stations
5: integrated data sources
GIS + time series: model of the real distribution network

Energy. The first step.

You have the data. We turn it into a system.

The right solution starts with understanding the operation.

Together, we map the available measurements, data sources and relationships between them. On that basis, we design an architecture and path that can begin with clear monitoring and, as benefits are verified, develop towards prediction and semi-autonomous control.

info@leris.cz