Hi all, I am Govinda CEO of SmartHelio.

We are really excited to launch 4 powerful analytics tools that can help solar utilities automatically predict and prevent solar plant failures and increase their overall performance.

We have taken the 4 most pressing challenges faced by the solar PV industry today (bad and missing data, inaccurate weather forecast, uneven & unknown panel degradation, and ineffective cleaning schedule) and designed easy-to-use tools with one click to solve these challenges. In the making of these tools, we have used our subject expertise - understanding of PV System Engineering, weather modeling, and climate change modeling, and coupled it with advanced computational methods and machine learning.

Please give your comments. Happy to answer your queries.

Tool [3/4]: Dynamic Cleaning Scheduler

Purpose of the Tool: The wide-area spread of the solar plants creates operational challenges for the solar asset managers. In most cases, the soil disposition across the plant is non-uniform and dynamic in nature. Moreover, meteorological events (like rainfall, snow, dust storm, etc.) play a vital role in the overall deposition of soil; hence, these factors should also be considered in the cleaning schedule. A fixed cleaning schedule that normally does not cover these factors, leads to inefficient cleaning and eventually leads to significant loss of energy production and resources (time, water, and human).

Creators: @kritz03 & @dorianguzman

The Methodology: The Dynamic Cleaning Schedule helps solar asset managers explore the most optimal timing for cleaning. This helps to maximize the energy production, reduce the cleaning cost and minimize the usage of water used for cleaning the solar panels.

Dynamic Cleaning Scheduler considers the historical & current soiling trends, meteorological & cleaning events, weather forecast, and the cost of cleaning and follows an AI-based optimization process to generate a dynamic cleaning schedule for different sections of the plant.

Tool [4/4]: PV Panel Analytics

Purpose of the Tool: Early degradation and permanent damage of the solar modules because of delayed interventions are becoming the major problems in the solar industry. As per one estimate, Solar PV waste will be one of the largest types of e-waste globally by 2050. With the module analytics service, we can remotely identify the degraded panels in the system and execute many more panel-level analytics in real-time with a click of a button on our computers. It also helps to optimize drone usage.

Creators: @sourabh470, divyansh & @humaid_mohammed_niyaz

The Methodology: Provides users with comprehensive monitoring and advanced diagnosis of PV modules. It analyzes the module level current, voltage, and temperature data collected by the Device. Post-processing of collected data informs the user about instantaneous power generation, energy generation on an hourly, daily, weekly and monthly basis. As PV module’s performance is sensitive to temperature change, the tool calculates the temperature coefficient of voltage as per IEC 60891 standard to assess the temperature sensitivity of different PV modules or different installations.

Tool [1/4]: AI-based Data Sanitization Service

Purpose of the Tool: The data acquisition system deployed in the solar industry frequently suffers from sudden communication failure and monitoring equipment malfunction. This leads to significant missing data sets and a lot of bad values in the system database. Therefore, any analysis based on these data sets leads to inaccurate results, higher assumptions, and bad conclusions.

Creators: @bhavya_dureja & @dorianguzman

The Methodology: The algorithm is formulated in a way that it first removes all the bad/corrupt data which are called as outliers using the PV system’s configuration and expected performance behaviour. Next, it identifies the data gaps which could be due to network/communication errors, bad weather data, or fault measurement devices. Finally, using AI & Machine Learning algorithms, the model first learns trends and seasonality of the data and later predicts these data gaps with high accuracy to reduce data loss.

Hello Everyone - I am Neeraj, CTO of SmartHelio. I take this opportunity to talk about the analytics tools. We would love to get some feedback and clarify your quarries.

Please feel free to add your suggestion (or quarries) in the comment section.

Tool [2/4]: Resource Forecasting Service

Purpose of the Tool: Variability and unpredictability of renewable resources (solar irradiance and wind speed) are the most pressing limiting factors of any solar/wind energy system. The current solutions lack localization of the forecast and they use only historical satellite data which is inaccurate

Creators: @rohan_karthikeyan

The Methodology: Provides users with comprehensive solar irradiance forecasts for any location in the world. It analyzes historical GHI data, learns trends and seasonality using deep/machine learning algorithms. It also takes into account climate change (IPCC RCP Scenarios), the microclimate of the location (Surface roughness, NDVI, Aerosol, water bodies, etc.), and human factors such as pollution, urbanization, etc. which helps us to achieve long-term forecasting accuracy up to 98%.