Click here to skip the introduction and jump into to the Data and Software Chapters.

In wastewater management, the synergy between software, hardware, and comprehensive services are essential components for utilities that aim to understand and optimize their wastewater ecosystem. In an age marked by environmental awareness, urban development and a commitment to resource efficiency, solutions like real-time wastewater intelligence can be a crucial value to achieve such an objective. Wastewater insights can support utilities in improving operational efficiency, workforce management, inflow of the wastewater treatment plant, and its image to attract a younger workforce generation. However, addressing these objectives and ensuring that the solution components work together is a challenging task, as it requires the input of various departments with different backgrounds, objectives, experience and outputs.

Kando's integration of AI-powered software, along with its specialized hardware and tailored services components, provide utilities a solution component combination that delivers in-depth insights into their wastewater quality that can lead to an optimal running wastewater ecosystem. Kando's API option enables the sharing of wastewater quality insights across different departmental platforms, allowing the data to be utilized utility-wide and support various objectives of different teams.

This article takes a deep dive into Kando wastewater intelligence solution, a Data as a Service (DaaS) platform, which integrates software, hardware, and expert services components. With a particular emphasis on discussing the software component, it explores how Kando leverages various wastewater and industrial data sets processed by its Artificial Intelligence (AI) and Machine Learning (ML) models to deliver real-time wastewater insights on the dashboard.

Kando’s wastewater intelligence solution automates the process of monitoring wastewater quality and sampling in real-time, keeping track of pollution, detecting wastewater quality deviations and patterns, and alerting the user in case of high-risk wastewater quality events via email and Telegram. Additionally, Kando has developed STREAMi, a virtual assistant designed to support utility pre-treatment and compliance teams in their daily operations. STREAMi uses natural language processing for easy access to wastewater insights, trends, and sampling. From identifying contamination events to offering strategic solutions, STREAMi simplifies complex data interactions and enhances decision-making.

The solution enables water utilities to:

Figure 1. Kando’s DaaS dashboard features & STREAMi

Kando wastewater intelligence solution is a comprehensive DaaS (Data as a Service) that seamlessly integrates software, hardware, and expert services, all powered by AI and ML technologies. This integrated approach enables Kando to deliver real-time wastewater intelligence through Kando’s User-friendly dashboard.

Figure 2: Kando Wastewater Intelligence DaaS is composed of Software, Services and Hardware

As depicted in Figure 2, Kando Wastewater Intelligence DaaS Solution is composed of three key components:

The system consists of the following attributes:

Kando uses multiple data sets to deliver real-time wastewater intelligence, categorized into four key pillars: Online Data, Utility Data, Historical Data, and Public Data. Each contributes equally to generating comprehensive insights on the wastewater quality in the collection system and its regulatory or operational impact.

Online Data, or Real-time data plays a critical role in enabling live insights, providing a snapshot of the current status of the wastewater system. This data is collected through Kando's IoT devices, the the A-EYE sensor (for more information, please view our hardware article), and processed together with its lab sampling data by a sub-ML model before being analyzed alongside the other data sets discussed below.

Kando's Historical Data consists of an extensive archive built over 11 years of R&D, which includes identified wastewater fingerprints, key parameters, and industry-specific processes for initial analysis of the type of event it is and its location.

To further improve accuracy, Kando integrates Utility Data to establish benchmark wastewater quality standards, GIS and permits.

Public Data is also utilized to help identify issues and predict future wastewater events.

These four datasets are continuously processed by Kando’s software, providing comprehensive real-time insights into wastewater quality status and deviations from wastewater regulatory thresholds and permits or its impact on the WWTP. Please find below a specific breakdown of each of the sub-datasets utilized.

Figure 3. illustrates the journey data takes through Kando’s wastewater intelligence system, beginning with the integration of various data types including public, utility, online, and historical data.

Before delving into the core functionalities of Kando’s software component and its data acquisition layer model (data value chain showcased in figure 4) of the previously mentioned data sets in chapter 4, it is crucial to understand the foundational technologies that empower this chain: Artificial Intelligence (AI) and Machine Learning (ML).

This section includes a brief overview of AI and ML, emphasizing how they differ yet complement each other in the world of data processing and analysis. Kando’s application of these technologies will be discussed, which is followed by a detailed description of the data flow and the extraction of actionable wastewater insights, facilitated by Kando’s AI and ML algorithms. This will introduce a comprehensive understanding of the integral role played by Kando’s data acquisition layer and how its processed by Kando's software to deliver real-time and predictive wastewater insights into Kando's dashboard.

Artificial Intelligence (AI) and Machine Learning (ML) are two interrelated technologies that play a crucial role in the development of intelligent systems, but it is important to note that they serve distinct functions within the broader spectrum of data analysis and interpretation.

AI is a wide-ranging branch of computer science concerned with building smart machines capable of performing tasks that typically require human intelligence. This encompasses a variety of activities, including problem-solving, pattern recognition, and understanding a language. AI aims to create systems that can function intelligently and independently.

ML, on the other hand, is a subset of AI focused specifically on the concept that machines can learn from data, identify patterns, and make decisions with minimal human intervention. ML algorithms allow computers to learn from and make predictions or decisions based on historical data. It is the method by which AI achieves its intelligence, the process through which AI learns.

The key difference lies in their scope and application: AI is the broader concept aimed at mimicking human cognitive functions, while ML is a specific approach within AI that teaches a machine how to learn from data, improving its accuracy over time as it processes more information. AI is about creating intelligence, whereas ML is about the learning process that supports AI's ability to adapt and improve.

To understand Kando's data integration flow that allows the solution to retrieve its insights, we will discuss figure 4 as depicted below displaying the data integration flow and the AI & ML models that are processing the data sets discussed in chapter 4.

The 4 sets of data are fed into Kando's advanced data acquisition system, where its AI model processes the data that captures and streamlines crucial details from the wastewater network ecosystem. This data is then refined by an AI-driven data cleaning process, ensuring it's free from errors and ready for analysis. The process continues with the application of ML models, such as anomaly detection and pattern recognition techniques, to develop a detailed quality event data model. This model excels at monitoring live sensor data, pattern recognition, pinpointing anomalies, and triggering automated sampling.

The value of these initial data insights is further enhanced through algorithmic analysis, leading to enhanced and comprehensive impact insights (discussed in chapter 5.2.1). These findings are then visualized on the Kando dashboard including notification to the user when a certain benchmark has been passed, providing a user-friendly platform for data interaction and decision-making.

To accommodate diverse utility needs and facilitate seamless integration into existing workflows, Kando also offers API access to its data insights, making its solution adaptable and easily incorporated into various wastewater management frameworks of different departments.

Figure 4. Data Flow

In the context of Kando's wastewater data flow, the applications of AI and Machine In the context of Kando's wastewater data flow, the applications of AI and Machine Learning (ML) are central to processing vast volumes of data and deriving actionable insights for its wastewater DaaS solution and API. ML and AI are utilized to tackle the complex challenges of understanding the multiple integrated data sets and allows the software component to analyze, and process wastewater data translated into insights from the complex wastewater network to optimize the network and operations.

Here’s how AI and ML contribute to each stage of Kando's data processing and the calculation of wastewater insights:

The integration of AI and ML within Kando’s dashboard provides a dynamic and interactive platform for clients to access real-time and predictive insights. ML algorithms drive the dashboard’s capability to present customized alerts, forecasts, and recommendations, in intuitive navigation and visualization of insights, enabling clients to make informed, data-based decisions quickly.

Figure 5. Example of an Event details card widget in Kando’s dashboard

Kando leverages a sophisticated data integration and analysis approach to provide actionable insights into wastewater quality events. To break it down, we will take metal coating factory insight detection as an example as shown in figure 5.

As discussed in previous paragraphs, and shown in figure 3, a load detection insight starts with the aggregation of four primary data sets: public data, Kando's extensive historical database (which includes unique fingerprint data), online date and utility-specific information. The latter encompasses GIS data, industry discharge records, factory permits, parameter benchmarks, and real-time wastewater data collected from the network.

This raw data combination is processed through Kando's advanced AI and ML models to transform this raw data into real-time, actionable wastewater insights. Initially, the data undergoes a series of algorithmic treatments—referred to as the data acquisition layer—to cleanse, organize, and commence the correlation process. This stage aims to align the data with specific criteria, facilitating early-stage models to utilize the pre-processed data for initial insights demonstrated in the dashboard insights.
​

Kando’s platform is built with robust cybersecurity practices to ensure the protection and integrity of utility data. This includes automated patch management, SSL encryption, secure communication across all components, and strict access controls. The system adheres to industry standards and best practices, aligning with frameworks such as ISO 27001 and SOC 2, and supports GDPR-compliant data handling where applicable. These safeguards ensure that sensitive operational and regulatory data remains protected at every stage, from field sensors to dashboard insights.

Kando's methodical and layered approach to data analysis ensures that clients receive a depth of insights that support informed decision-making in wastewater management. Through its dashboard, Kando offers a great overview into the intricacies of the wastewater quality dynamics within the collection system, enabling utilities to optimize operations, improve environmental compliance, and enhance overall sustainability efforts while still meeting the community demands.

The following sub chapters will discuss the different types of features available in Kando’s dashboard.

Kando's dashboard offers a comprehensive assessment of organic events contribution across the entire sewer network. It effectively highlights key inlets, sub-areas, and factories, directing users' attention to critical points for optimized management. The monitoring process is initiated at the WWTP and its inlets, then progressively extends upstream, leveraging real-time data collection and sophisticated assessment algorithms as discussed in chapter 4. This strategic approach ensures targeted and efficient monitoring, enabling users to identify and address issues proactively.

The "Potential Savings" feature serves as a powerful motivator for reducing the organic load, by highlighting the economic benefits achievable through optimization. This feature calculates potential savings by integrating key operational parameters of the wastewater treatment plant, including organic events specific to the utility.

The foundation for these calculations is either a comprehensive composite sampling conducted at the outset of the implementation phase, or values directly provided by the utility. By presenting a clear correlation between reduced organic events and financial savings, this feature encourages proactive measures to enhance WWTP operations, offering a tangible metric to assess the monetary impact of optimization efforts.

The "Organic Event Status" is designed to serve as a critical indicator of the organic load at the wastewater treatment plant (WWTP), providing a direct reflection of the current state compared to baseline levels established at the initiation of the implementation. This feature not only quantifies the organic load, but also illustrates the temporal progress, offering users a visual representation of changes over time. By comparing current data to the start of Kando’s implementation, stakeholders can accurately measure the effectiveness of its deployment to reduce the organic events, enabling informed decision-making to further optimize WWTP operations.

The "Organic Trends feature" on Kando's dashboarddemonstratesthe daily fluctuations in organic events, providing an in-depth analysis based on the severity and duration of these occurrences. This analytical tool visualizesthe trajectory of organic loads within a selected period, enabling users to understand patterns and anomalies with precision. By delivering a nuanced view of daily event scores, the featureuncovers underlying trends that could inform strategic decisions, facilitating proactive adjustments to wastewater treatment protocols. This capability to track and analyze temporal trends is instrumental in fostering an adaptive management approach, allowing for the optimization of processes and the efficient allocation of resources to mitigate the impact of organic events on the WWTP operations.

The "Inlet Contribution feature" on Kando's dashboard demonstrates the most impactful inlet, thereby prioritizing the wastewater management efforts. The feature presents a bar chart that maps out the event contributions from each inlet alongside their respective flow ratios. This visualization facilitates the prioritization of inlets based on their organic event contributions, allowing stakeholders to strategize or refocus their management and intervention strategies where they are needed most. Employing this targeted approach enables efficient resource allocation, enhancing the efficacy of wastewater treatment operations by concentrating on areas with the greatest potential for optimization and impact.

The "Organic Event Pattern feature”on Kando's dashboard demonstratesthepatterns of organic events, a foundational step for enacting timely and effective operational actions at the wastewater treatment plant (WWTP). This feature displays a heatmap to visually represent the severity across the day, providing a clear, at-a-glance understanding of when and how significantly organic events fluctuations occur. By delineating these patterns, the feature facilitates strategic planning and immediate response to mitigate the impact of organic event peaks, thereby optimizing the operational efficiency and overall performance of the WWTP. This level of insight is instrumental in proactive wastewater management, ensuring that interventions are both timely and precisely targeted to meet the plant's needs.

The "Sampling feature” on Kando's dashboard demonstrates the visualization of laboratory results obtained from the wastewater treatment plant. It presents users with an intuitive interface and visualizes the lab results together with a comprehensible insight. By categorizing these results into a chart based on parameter type or sampling date, the feature allows for an efficient review and analysis process. This visualization empowers users to easily interpret complex lab data, facilitating quick and informed decision-making regarding the plant's operational adjustments and health. Whether focusing on specific contaminants or monitoring changes over time, the Sampling feature ensures that critical laboratory insights are accessible and actionable, significantly contributing to the optimization of wastewater treatment strategies.

The Time Selector feature on Kando's dashboardprovides flexibility to navigate through different time periods to uncover specific insights. This capability ensures that stakeholders can conduct temporal comparisons and trend analyses with ease, making it a vital tool for strategic planning and operational optimization in wastewater management.

Integral to understanding the dynamics of the wastewater treatment plant (WWTP) or city network, the "Map" that is found next to the dashboard in a different tab of the dashboard demonstrates the assessment and real-time information concerning the organic event sources. Users can interact with the map by clicking on the WWTP or specific areas, which then unveils an assessment pie chart detailing the contributions to the organic load from downstream areas or factories. This interactive feature enhances the utility's grasp of the organic load distribution and pinpoints the significant contributors. By offering a spatial analysis complemented by detailed pie charts, the Map facilitates a deeper understanding of the organic load dynamics, enabling targeted interventions and informed decision-making that will allow the team to improve wastewater quality in the collection system.

The "Left Side Panel" feature on Kando’s dashboard is a strategic tool designed to spotlight the most significant contributors to high-risk wastewater events (pollution) within the network—whether they are inlets, areas, or factories. It is organized by a calculated score that evaluates the impact of each monitored or assessed item on the overall organic load. An icon beside each entry clearly denotes whether the location is under active monitoring or has been assessed based on available data.

The "Pie Chart feature " in the dashboard is designed to visualize the organic event contributions from different sources within the wastewater network, such as inlets, sub-areas, or factories. By employing a pie chart display, this feature calculates and presents the percentage contribution of each entity, offering a clear and concise overview of how different sections contribute to the overall organic load. The values represented in the pie chart are derived from a detailed assessment of the contribution scores for each factory, enabling users to easily comprehend the distribution of organic load contributions. This facilitates informed decision-making by highlighting areas that require targeted intervention for optimized wastewater management.