PVPredict develops machine learning algorithms for advanced statistical performance monitoring and PV generation forecasting. 5 years in development with support from the Israeli ministry of energy has led to a SaaS product now available to monitoring companies to enable their customers to effectively ascertain the health of their systems every day while enabling grid managers to reclaim control of the grid flow.

Our target customers:

PV monitoring companies
PV fleet owners
Inverter manufacturers
Energy traders with PV in their portfolios
Transmission System Operators
Distribution System Operators


Over half of the installed PV power in most countries is installed and managed by residential or commercial customers.
These systems lack good performance monitoring due to the cost of hardware and software.
These customers are not aware of their system's availability beyond the receipt of their monthly electrical bill.
Yet they aggregate into large virtual PV power plants that feed the national grid as an unknown variable to the grid manager.
So the system owner has no idea how his system is performing from an objective point of view, if at all and the entities manageing the grid at both the national and local level are dealing with rogue power plants they cannot take into account in the delicate affair of grid management


Statistical performance monitoring system that requires:

No hardware
No input of system configuration
No analysis software

Only requiring:

Inverter or generation meter output
GIS location

Benefactors of our system include:

Owners of small, medium even large PV systems
PV monitoring companies
PV fleet owners
Inverter manufacturers
Energy traders including PV in their portfolios
Transmission System Operators (TSO)
Distribution System Operators (DSO)


PVPredict has developed machine learning algorithms that predict the next day’s hourly generation of any solar PV inverter without the use of sensors or knowledge of the system configuration.
The algorithms also show promise in predicting failures before power loss occurs.
These algorithms are the core technology for the following services:

State of Health report every morning for the owners of small solar PV systems with no performance monitoring capabilities – a useful customer retention product for power selling utilities

Early warning of impending system faults

Accurate day ahead and hour ahead generation predictions for energy traders with no necessity for on site sensors

Reducing spinning reserve by making available to the Transmission System Operator (TSO) manager each neighborhood with installed PV as a virtual power plant with next day’s hourly generation delivered every evening

Managing voltage and current on the low voltage distribution grid managed by the Distribution System Operator (DSO)


Mike Green

Electrical engineer; Owner of M.G. Lightning ltd. design and consulting; former CTO of Arava Power; consultant to the Israel Standards Institute on rooftop solar; represents Israel in the IEA-PVPS Task 13  researching the efficiency and reliability of PV systems

Eyal Brill Ph.D 

Expert on unsupervised machine learning; owner of Decision Makers ltd.; post-doctoral degree from the University of Maryland; Deputy Head of the MOT faculty at the Holon Institute of Technology in Israel

Adam Hirsch Ph.D
Business Development

Former NREL researcher with energy efficiency and smart grid experience. Lecturer in Energy Systems and Sustainability at the Herzliya Interdisciplinary Center. Bachelor’s (Harvard University) and Ph.D. (University of California at Irvine) degrees in the Geosciences.

Shimshon Rapaport
Electrical Engineer

Experienced in PV system design, simulation and solar energy research, provides hardware/software integration, statistical analysis and web development. BA in mathematics and BSc in electrical engineering from the University of Colorado Boulder.

Adi Brill
Software Developer

Barak Brill
PhD Student, Programmer

Igor Lukatski
Electrical Engineer

Senior PV and electrical system designer, advanced PV system simulations and on site installation and supervision. BSc in electrical engineering from Afeka college of engineering, Tel Aviv.

Michael Britvin,
Data Integrator



is the home of the SolarEra.net backed program for our Distributed Energy Resource Management System (DERMS) pilot project designed to enable a DSO to manage the voltage in residential distribution grids with a high level of solar PV penetration while aggregating the neighborhood into a virtual PV plant for reducing spinning reserve.

Ministry of Energy

Long time supporters, three rounds of investment in our technology

IEA PVPS Programme

A member of Task 13 since 2010, we have collaborated with PV experts from over 25 countries. our development can be found described in the 2017 report “Improving Efficiency of PV Systems Using Statistical Performance Monitoring”.

DecisionMakers ltd.

DecisionMakers ltd. is the company owned and managed by Dr. Eyal Brill in which the machine learning algorithms driving PVPredict where developed.

M.G. Lightning

Owned and managed by Mike Green is the vehicle that drove the first 5 years of Pvpredict development opposite the world solar PV market.


Greeneum incentivizes and optimizes renewable energy trading using blockchain and personalized AI


Jan 2013

Received Min. of Energy “Heznek” grant to develop machine learning algorithms for predicting next day’s hourly PV generation without the use of sensors

April 2014

Our algorithms proved more accurate than those used by commercial companies in Europe for predicting next day’s hourly generation for selling on “Next Day Market”

Dec 2014

Successful conclusion of Min. of Energy development grant, begin preparing marketing foundation for selling SaaS to residential customers as a statistical performance monitoring solution requiring no sensors or input from the owner

Jan 2015

Received 2nd Min. of Energy “Heznek” grant to develop machine learning algorithms that will predict faults before they occur using clustering regression-tree statistical methodology aided by expert input

Sept 2015

Received Min. of Economics (Innovation) grant for joint development with University of Cyprus of algorithms for the early and reliable detection of degradation in Photovoltaics (PID)

March 2016

Our server platform is capable of delivering “State of Health” mail to PV owners each morning as well as warning immediately on loss of power and loss of communication

Dec 2016

Successful conclusion of our “Early fault warning” development funded by Min. of Energy

Jan 2017

Received SOLAR-ERA.NET project as CONSORTIUM LEADER in a PILOT project in Cyprus and a Kibbutz in Israel to apply our algorithms for managing the DISTRIBUTION OF RESIDENTIAL PV ENERGY IN AN LV DISTRIBUTION GRID (Utility market)

Oct 2017

Began marketing actively to the utility support market with a presence at the European Utilities Week in Amsterdam

June 2018

Integrated with the Greeneum – the blockchain-based platform for renewable energy, supplying the SolarPET.

Dec 2018

Completed advanced algorithms for use with hour ahead and 5 minute ahead energy forecasting in Virtual Power Plants, energy trading and for continuing the DERMS research.

Jan 2019

Pilot with FSIGHT at Maaleh Gilboa DERMS project, supplying the residential PV predictions

March 2019

Pilot project with Emulsionen, a Swedish monitoring company interested in offering their Prosumers the PVpredict performance monitoring capability.

July 2019

Pilot with Kyocera to enable predictions in Virtual Power Plants.

Sep 2019

Pilot with Repom, Romanian based renewable energy asset manager for supplying day ahead predictions for utility grade dispatched PV power stations



You can call or WhatsApp us at
+972 54-499-9169


Email us at: