top of page

M A N O H A R

Cassandra

Smart effluent monitoring system for breweries

"Increased the annual revenue by 15% and brought down the brewing time from 5.33 hours to 4 hours, achieving a 25% reduction."

About the project

3 Months

Duration

My Role

Product Designer

Sun King Brewery, the largest brewery in Indianapolis, was experiencing issues with their brewing process and required a tailored pH monitoring system for the waste generated.

Company

Endress Hauser

Team

2 UX Researchers 1 Product Designer

Problem Statement

How might we develop a sustainable, preemptive and digital system that enhances the production capacity, utilizing effluent monitoring?

Secondary Research

Understanding the current brewing process through desk research and field visit

We conducted an online research study and visited Sun King Brewery in Indianapolis to gain a fundamental understanding of their current effluent monitoring process.

Key findings from desk research and field visit

01

The Beer brewing process consists of five distinct tanks.

02

Effluent is generated at every tank.

03

pH level is measured before discharging the effluent.

04

Brewers manually collect samples for testing.

05

Nine values are collected from samples.

06

If any value crosses threshold, the whole batch is discarded.

Primary Research

Exploring the brewery experience: Insights from interviews with key personnel

We conducted interviews with the lead brewer, maintenance director, and quality control manager to get an understanding of how they interact with the equipment and what are the difficulties faced by them.

Key findings from interviews

QC manager spends 30-60 min daily collecting test samples.

pH levels must be within specific threshold per batch.

Brewers detect anomalies in samples.

Data is stored on hard drives.

Anomalies are notified via email only.

Brewery maintains FDA/EPA effluent reports.

Quality Control Manager | Sun king Brewery

"pH value is the main component to establish the consistency. It is manually measured and updated in the excel sheet daily."

Maintenance Director | Sun king Brewery

"We have to align with FDA and EPA regulations, the effluent generated has to meet the set conditions before we discharge it."

Analysis

Using affinity mapping to analyze collected data and derive a problem statement

By employing affinity mapping, the team categorized and examined pain points and challenges encountered throughout the brewing process. Subsequently, the gathered data points were analyzed to formulate a problem statement based on the identified issues.

Challenges faced by users

Inefficient Sample Collection

Brewers rely on manual collection of samples for pH testing, a time-intensive process that not only slows down production but also introduces opportunities for human error.

Excessive Time Commitment

‍The Quality Control manager dedicates 30 to 60 minutes daily to collecting test samples, this diverts valuable resources away from other critical quality assurance and optimization activities.

Risk of Batch Loss

If pH levels exceed the defined thresholds, the entire batch must be discarded. This results in significant material waste and financial loss, directly impacting the brewery's profitability.

Primitive Data Management

Data from pH tests is stored on physical hard drives and shared via email, a method that lacks real-time accessibility and scalability.

Ideation

Improving brewing efficiency with E+H sensors: Focusing on user pain points

The team initiated brainstorming sessions to explore ways of enhancing the brewing process's efficiency through the implementation of E+H sensors. They carefully analyzed the pain points uncovered during the research phase and prioritized the requirements and desires of the target users throughout the ideation process.

Various conceptual solutions were developed to address the identified pain points

Solution 1

Sensor collects and transmits data

Data is received and stored in the cloud

The user receives the values on their device

Step 1

Step 2

Step 3

Solution 2

Two sensors collect and transmits data

Data is received and stored in the cloud

The user receives the values on their device

Step 1

Step 2

Step 3

Two sensors collect and transmits data

Data is received and stored in the cloud

New readings are sent to theAI

Step 1

Step 2

Step 3

AI analyzes the data and predicts the pH levels

Step 4

Step 5

The user receives the values on their device

Solution 3

🌟 Client Approved Solution

Initial Client Feedback

The brewers at Sun King evaluated and assessed the proposed concept solutions

The three concept solutions were showcased to the Sun King brewers, who were requested to assess each solution across four key factors: Productivity, Efficiency, Cost, and Desirability. Using a scale from 1 to 5, with 1 indicating the lowest and 5 representing the highest rating, the brewers provided their evaluations for each factor.

Solution

Applying predictive analysis to forecast and regulate pH levels for effective monitoring of effluent

After receiving input from the client, we have gained a comprehensive understanding of their emphasis on maximizing efficiency and productivity. Taking into account the requirements of both users and stakeholders, our team presented a product proposal that leverages predictive analysis to anticipate pH levels in each tank's effluent and proactively detect anomalies in advance.

With predictive analysis, anomalies are detected in advance.

Data is stored and backed up in the cloud automatically.

The sensor collects 7 different values from the sample including pH level.

Data can be easily downloaded to maintain required reports.

Paper Prototypes 

With predictive analysis, anomalies are detected in advance.

Prototype 1

Prototype 2

Prototype 3

Prototype 4

High Fidelity Prototypes

Presenting Cassandra

Comprehensive dashboard that shows 8 crucial values with their respective measuring unit

The sensor not only records pH values but also captures other crucial data points significant for brewers. These additional values, along with their corresponding dates, are showcased on a unified dashboard, allowing brewers to access and monitor all the essential information from a single screen.

228_3x_shots_so.png

User-friendly design for streamlined understanding

The sensor data was presented in a way that ensured a clear comprehension, with special attention given to accommodating brewers limited experience with digital systems, mainly due to their extensive reliance on manual methods.

791_3x_shots_so.png

Efficient and simplified data management for reporting needs

The data concerning various brews are readily accessible and can be easily viewed and downloaded to facilitate the creation and distribution of reports.

497_3x_shots_so.png

Providing transparency with visual representations for system's prediction accuracy

The prediction accuracy is visually represented using charts, offering a clear and transparent way to showcase the data. Additionally, the charts provide numerical details that highlight the accuracy of the predictions.

714_3x_shots_so.png

Instant Notifications

Once the system detects that the pH value has exceeded the threshold, it immediately sends a notification to the user through the in-app notification system.

672_3x_shots_so.png
gauge-svgrepo-com (1).png

Testing Process

Participants: Recruit participants who align with the target audience, such as quality control managers or technical operators.

Tasks:

  1. Navigate to the "Threshold" screen and set specific values for pH Liquor Tank, Chloride, and Hardness.

  2. Locate a specific brew (e.g., "Wee Mac") and view its details in the Dashboard.

  3. Interpret the pH Deflection graph and describe what the warning icons mean.

  4. Export data from the "Dashboard" screen.
     

Facilitation: Ask participants to verbalize their thought processes as they complete tasks. Observe and note any hesitation or confusion.
Outcome: Gather qualitative data on usability issues and areas of improvement, such as field validation, icon clarity, or navigation flow.

Usability Testing

Moderated Usability Testing: I conducted in-depth moderated usability testing to evaluate the Cassandra platform's dashboard, threshold settings, brew management, and analytics screens. Participants completed tasks such as setting thresholds, interpreting pH analytics, and navigating between brews. Observations revealed opportunities to enhance navigation flow and visual clarity, leading to targeted design improvements.
 

Unmoderated Usability Testing: To validate the platform's usability in real-world scenarios, I conducted remote unmoderated testing. Users independently completed tasks like adjusting thresholds, interpreting warning notifications, and analyzing prediction metrics. Insights from these tests led to the addition of tooltips and improved iconography, enhancing the overall user experience.

gauge-svgrepo-com (1).png

Testing Process

Participants: Invite participants to access the system remotely and provide task instructions via email or a usability testing tool like Maze or UserTesting.
 

Tasks:

  1. Update the threshold for "Sulphate" and "Free Chlorine" in the "Set Threshold" screen.

  2. Navigate to the "Brews" section and locate a specific brew.

  3. Interpret the warning message displayed on the lock screen notification.

  4. View and analyze pH Analytics to identify tanks exceeding the set threshold.
     

Data Collection:

  • Use click-path analysis to monitor navigation patterns.

  • Record task completion times and success rates.

  • Provide a post-task survey to gather user feedback on clarity and usability.
     

Outcome: Identify usability issues, such as unclear icons or hard-to-find features, and validate design decisions like the effectiveness of visual cues and graph interpretations.

Moderated Usability Testing

Objective: To observe and understand how users interact with the dashboard, prediction analytics, threshold settings, and other features in a controlled environment.

Key Components Tested:

  1. Dashboard and Metrics Overview:

    • Assess if users can quickly understand and interpret the prediction accuracy metrics (e.g., Mesh Tank, Boiling Tank) and status indicators.

    • Test how easily users navigate between sections like "Dashboard," "Predictions," and "Accuracy."

  2. Threshold Settings:

    • Observe if users can set and adjust thresholds for parameters (e.g., pH levels, chlorine) without confusion.

    • Check whether the labeling and input fields are intuitive and accessible.

  3. Brews Management:

    • Evaluate if users can add and manage new brews effectively.

    • Observe if users understand the visual organization of brew names and can locate specific brews quickly.

  4. pH Analytics and Alerts:

    • Test if users can interpret pH analytics, including the deflection graph and the warning system for exceeding thresholds.

    • Observe how users react to and interpret notifications about anomalies

Unmoderated Usability Testing

Objective: To evaluate how well users can independently navigate and complete tasks on the Cassandra platform in real-world conditions.

Key Components Tested:

  1. Navigation and Accessibility:

    • Test if users can independently locate key sections such as "Brews" and "Accuracy" without external guidance.

  2. pH Analytics and Alerts:

    • Validate if users can interpret pH predictions and warnings accurately when not guided by a moderator.

  3. Threshold Settings:

    • Assess if users can adjust thresholds correctly without running into validation errors or unclear feedback.

  4. Dashboard Usability:

    • Test if users can interpret prediction accuracy metrics and statuses without confusion

Measuring Success

Impact made by Cassandra

Saving Resources

With pH prediction, brewers can prevent wastage by identifying anomalies early and taking corrective action.

Saving Resources

The system records every brew and generates reports in the required format, saving brewers time.

Increased Efficiency

The sensor have automated the collection of required values from samples, which was previously done manually by brewers.

bottom of page