SmartWater: an intelligent software for monitoring at real-time water networks (2011-2017)
with Sense4Green

Aging infrastructures are responsible worldwide for remarkable water dispersion, high operational, maintenance and energy costs


SmartWater is an intelligent software, which is designed to improve the operational efficiency of water infrastructures, support decisions and maintenance planning, and reduce water and energy consumption.


Through the on-line analysis of heterogeneous sensor data from the network and unstructured data, it accurately detects anomalous events at real-time such as water leaks, transients, and anomalous trends that can lead to a critical state and inefficiencies

SmartWater is innovative also for scaling to very large networks, being robust, accurate, and economically sustainable.


It was successfully applied to: 

  1. a water utility serving 5000 people affected by high water dispersion;

  2. a small district to monitor water quality.

Early-warning and climate-change mitigation in drainage systems (2014)
with Marco Maglionico at University Bologna and KIC-Climate

Monitoring water flow and level data is critical to control the pollutant level of the sewer system in case of heavy precipitations since the system transports both stormwater runoff and sewage.


SmartWater algorithms were applied to analyze data collected from flow and water level sensors deployed in a combined sewer system in Italy, and shown their benefits in enhancing system control and mitigating effects caused by unexpected climate events

On-line smart metering analysis (2013)
with Sense4Green

I designed an on-line smart meter analytics system that scales to large volume of data and devices for relying on a distributed intelligence approach built in an innovative hierarchical model-based manner. It monitors smart meter data at individual and aggregate level, detects data patterns, and forecasts data.


The system can be applied to monitor loads in a large area at different aggregate level, or to enhance control in a microgrid and maximize its local renewable sources, or to dynamically reconfigure microgrids when switching to islanding mode. 

SmartPV: an on-line IoT monitoring for enhancing the reliability and performance of PV plants (2013)
with Sense4Green

SmartPV is an innovative sensor-based distributed control system for monitoring at real-time PV plants to improve their reliability and power efficiency, and to better integrate its production with the grid.


The system is able to monitor electrical/physical parameters, and analyze heterogeneous data at real-time to detect panel failures, power losses, and better predict power generation.

Demand response management in microgrids via real-time fine-grained monitoring and forecasting​ (2013)
with Sense4Green

SmartMicrogrid is an innovative system for monitoring at real-time and predicting the request and production within an electric microgrid. More precisely, the system monitors at real-time and with fine granularity all of the local electric loads, the energy produced by local renewable energy sources, and the energy stored. It predicts the amount of energy needed from the grid, and dynamically reschedules loads if a peak is approaching.


Advantages of this approach:

  1. Enhanced control and forecasting of the energy needs over the microgrid

  2. Better exploits the use of renewable energy sources by shifting loads and storage, thus reducing energy costs and carbon dioxide emissions.

SmartEnv: a real-time automatic monitoring sysem with quality guarantees (2006-2013)
with MIT and Sense4Green 

Wireless senor networks are known for their potential to remotely observe the physical world at high resolution and reasonable costs. However, sensor faults, communication disruptions, and unexpected events can affect the quality of the observations.

Moreover, the complexity of analyzing at real-time the huge volume od data produced at sensors often hampers its potential.


These problems have been addressed by SmartEnv, an innovative WSN middleware system for monitoring at real-time physical/electrical attributes with quality guarantees. The system is able to answer queries, predict values, detect/diagnose data anomalies at real-time, and detect data properties such as trends and temporal-spatial correlations.


SmartEnv is innovative with respect to current systems for:

  1. employing distributed adaptive intelligence,

  2. providing guarantees on the quality of the observations even in the face of temporal communication disruptions, sensor faults, and high data instability,

  3. providing on-line data analytics and distinguishing between faults and external events,

  4. being dynamically adaptable. Preliminary experiments performed over real sensor data (e.g., temperature, pressure) have indicated its effectiveness. ​

In addition, SmartEnv can assist sensor deployment in computing the number/position of sensors to be deployed in order to get probabilistic guarantees on the quality of the future service. 

Disruption-tolerant underwater sensor monitoring (2009)
with JRC 

Underwater sensor networks have the potential to enhance our ability of observing underwater physical phenomena, warning against natural disasters, and protecting critical areas. However, the adverse underwater conditions and the limitations of underwater acoustic communication such as its volatile link quality remarkably affect the quality of monitoring.


We analyzed this problem and proposed a model-based on-line monitoring scheme that provides probabilistic guarantees in the presence of transient communication disruptions.