ABOUT WEOTech LTD.
WEOTech LTD., which operates under as a registered company in Province of Manitoba, develops and distributes the Proprietary software applications for Waste Water Process Optimization purposes. According to the plan, the commencement of operations was scheduled on November, 2020. Our solution is estimated to reduce energy costs by 16.7% and alarms by 97%. Besides, the reduction of Carbon footprint will be in alignment with the Energy optimization, reduction in water &, and organic waste delivery schedules. Also, we provide a solution to save costs for the facilities by reducing the human intervention aspect of monitoring and attending to pumps, monitors, and alarm. The Data-driven solution is intended to convert, idle data, into a positive cash saving solution for the Wastewater Treatment companies in Manitoba, and to other Provinces in Canada.
BUSINESS DESCRIPTION
The Wastewater Treatment Industry is a mixture of various intricate series of biological, physical, and chemical processes that are used to treat and remove contaminants or pollutants from wastewater or sewage. Water treatment represents the most significant energy use for most municipal governments. There are four segments in Water Treatment Industry, comprising Wastewater Treatment Plants (WWTP), Drinking Water Treatment Plants (DWTP), Wastewater Pumping stations (WW Pumping), and Drinking Water Pumping stations (DW Pumping). There are many chances to reduce energy utilization, lower electrical peak needs, and decrease Greenhouse Gas (GHG) emissions in water treatment sectors. The effective operation of wastewater treatment plants (WWTPs) is vital to make sure a sustainable and friendly green environment is maintained. The Co2 emissions arising from water and wastewater transport and sub-optimal treatment is also a concern. An analysis of electric energy intake, specifically, shows the most dominant energy end-utilization in the water treatment sectors is pumping, signifying 1.9 TWh, with 65% of all energy use.
WEOTech Solution
WEOTech solution to this challenge will enable a decrease in energy use and energy costs; Lower peak demand, and Diminish Greenhouse Gas (GHG) emissions within the context of the Waste Water Treatment Sector. Our proposed predictive control can anticipate the incoming Wastewater intake rate (WWIR) and to adjust the reservoir buffer accordingly. We are introducing a two-stage framework using ML, Learning Stage, and Operational Stage.
Learning Stage
We train the model on an emulated environment, constructed with supervised learning (SL - initial learning stage). The average WWIR primarily relies on the period of the year or day. Initially, during the first stage, initial learning of the control policy takes place. Forecasting WWIR based on fuzzy logic is the first step.
TIME series are observations recorded sequentially over time, such as Environmental Sensor Data, and historical Wastewater Intake Rate (WWIR). They are an important subcategory of data streams in which the data is not only temporally ordered, but the exact time of observation is also recorded (explicitly or implicitly). As mentioned before, we need to forecast the amount of optimal WWIR Forecasting time series is a significant machine learning problem. WWIR forecasting is effective in WW and drought management. For forecasting WWIR regression, we used Fuzzy logic. The term fuzzy refers to things that are not clear or are vague, and “fuzzy logic is an approach to computing based on "degrees of truth" rather than the usual "true or false."
We use the adaptive neuro-complex-fuzzy inference system (ANCFIS) architecture;
For general multivariate time series forecasting, multiple-input-single-output (MISO) networks are used.
Previous Data containing information about the pump's active power and frequency, the wastewater tank /reservoir level, WWIR, and outflow rate are collected from the SCADA system of the WWPS This historical data is used for:
(i) WWIR probabilistic forecasts;
(ii) Create data-driven models that emulate the physical environment;
(iii) Set up episodes are combining historical data and forecasts.
A state vector is constructed from the environment using data such as reservoir level, Wastewater Intake Rate (WWIR) forecasts, pumps online, and current operational set-point, acquired by a set of sensors (IoT) placed with SCADA and subsequently fed to the predictive control strategy. With this information, the algorithm selects actions and power set-point for each pump unit.
Operational stage
The trained data is transferred to an operational setup and applied to the physical wastewater pumping station. ML learning falls under supervised, unsupervised, and reinforcement. We intend to use Reinforcement Learning (RL) at this stage, which is one of three basic machine learning paradigms. Reinforcement learning (RL) is a field of machine learning involved with how software drivers take up measures in an environment to maximize some notion of collective reward and minimizing its penalty. We apply data-mining algorithms to create models from data and generate synthetic data for pre-training the RL algorithm, without the need to interact with the system physically. Reinforcement learning differs from other machine learning methodologies in that the algorithm is not instructed on how to execute a task but works through the challenge on its own. Probabilistic forecasts generated through the learning of the WWIR are used as one of the inputs of the RL control algorithm, to provide the control strategy for uncertainty.
WEOTech History
We build the real value!
March 2020
Idea and Business Plan
May 2020
Co-operation with Manitoba Technology Acceleration (MTA)
November 2020
Company Foundation
March 2021
Web Application Development Planning
October 2021
Patent Registration
August 2022
Web application Phase 1 – Learning Stages
September 2022
Phase 2 – Operational Stages – Lab Production Planning
WEOTech LTD Team
Meet the people who can make it happen
Azadeh Shirvani Boroujeni
Chief Executive Officer
Mohammad Hassan Alizadeh Nazmi
Director of Operations
Saeed Noroozi
Chief Technology Officer
Morteza Mollazainali
General Manager