Posted on

Erika’s Status Report for 04/12/2025

Progress this Week:
This week, I redesigned the model house structure to incorporate finger joints. After observing structural weaknesses during our interim demo, I realized that the original flat-edge design was not sturdy enough, especially when using acrylic sheets for the inner walls that lacked sufficient support. I also met with my team to finalize where we want to include the electronic components so I can include holes for those wire feedthroughs. To address this, I modified the CAD models for all the walls to include finger joints and feedthrough holes for better structural integrity and easy integration. I then regenerated the corresponding DXF files for laser cutting.

Following these updates, I re-laser cut most of the acrylic and wood sheets. This process included completing the full assembly in CAD to ensure the finger joints fit together cleanly.

Challenges
My progress is currently on schedule. While redesigning the house model and re-cutting parts was not initially planned, this adjustment was necessary for the long-term stability and success of our prototype. I have built in buffer time for tasks like this, so I’m confident we are still aligned with our project timeline.

We also placed a lot of orders this past Wednesday, so I hope those arrive early next week so we can continue wrapping up the final pieces of our project.

Next Steps:

  • Finish laser cutting the remaining few acrylic sheets

  • Assemble the redesigned model house

  • Begin integrating the LLM chatbot using the Ollama setup that Maya worked on

  • Set up wiring and begin coding the LED ring to simulate the passage of time in our simulated reality environment (where 1 simulated day = 5 minutes in real time)

Posted on

Anya’s Status Report for 04/12

Accomplishments

  • Integrated real-time sensor polling and displayed it on the dashboard

  •  Developed a responsive FastAPI + HTMX dashboard to display live sensor values

  • Implemented Chart.js charts that auto-refresh every few seconds without full page reloads

  •  Latency about 10s for the API calls, page refreshes every 30s

Progress

Reduced latency by introducing in-memory caching for the a CNN solar forecast endpoint

  •  Forecast results are now cached and refreshed every 10 minutes, eliminating redundant model retraining on every request

  • Slightly behind progress, need to integrate all the charts in the same app with routing

  • Verification

Performed basic verification of the CNN forecasting model to ensure it meets the intended design requirements for solar prediction accuracy.

  • Input Feature Sanity Checks:
    Verified that time-series input includes correct power, sin_hour, and cos_hour features, resampled to 15-minute intervals.
    Resampling was validated via debug logging (len(df), delta distributions).

  • Window + Horizon Coverage:
    Model was trained on sliding windows (e.g., 48 steps = 12 hours) with a 24-step forecast horizon. These parameters were confirmed to provide full diurnal coverage and enough context for trend learning.

  • Envelope Alignment Verification:
    The daylight envelope was implemented and centered around peak solar generation (13:00–14:00) to prevent unrealistic output at midnight. Forecasts before and after masking were compared to confirm correction.

  • Polling Interval Consistency:
    The system pulls data using HA/api/history/period endpoint. I confirmed that sensors such as sensor.maya_solar_panel, sensor.maya_fan_power, sensor.maya_fan_voltage, and sensor.maya_fan_current are polled with sufficient granularity — typically every 5–10 minutes.

  • Each response is parsed to validate:

    • state is numeric and finite

    • last_changed timestamps are increasing and consistent

    • There are no large gaps in time , > 1 hour)

  • Resampling Debug Logs:
    Added internal logging to capture the number of raw and resampled datapoints. This ensures the model always receives enough sequential history for the sliding window.
    Used interpolation (.interpolate()) on the resampled time-series to ensure continuity in data, even if occasional values are missing from Home Assistant.

  • latency tests via time curl http://localhost:5051/api/endpoint to make sure the dashboard refresh updates meet design and usecase requirements

Next Steps

  • Test LP behavior under forecast uncertainty

    • Run LP with low vs. high solar predictions

    • Assess how schedules shift and whether they violate any constraints

  • Stress-test LP under edge-case constraints

    • Zero solar availability

    • Multiple high-power devices requiring overlap

    • Very short vs. very long runtime constraints

Test LP reactivity to live data

  • Feed in updated solar forecast every hour

  • Ensure LP returns updated schedules quickly (<2 sec)

 

Posted on

Erika’s Status Report for 03/08/2025

1. Accomplishments This Week
This week, I focused on two key tasks:

  • Design Review Report: I collaborated with my team to write and refine the design review report, ensuring we clearly outlined our project scope, system architecture, and implementation plan.
  • CAD Model for Demo House: I continued developing the CAD model for our demo house. Specifically, I designed a two-layer wall structure featuring wood on the exterior and acrylic on the interior. This design allows us to sandwich wires and electronic components between the layers for a cleaner, more organized prototype. I also worked on integrating component placements within the model to facilitate efficient wiring and sensor integration.

2. Progress Status
I am on track with my individual tasks for the week. The design review report was completed on schedule, and progress on the CAD model is aligned with our timeline. However, I will need to start laser cutting soon to ensure I do not run into any issues later on.

3. Next Week’s Deliverables
Next week, I plan to:

  • Finalize the CAD of the demo house and determine kerf to verify fit.
  • Laser cut the wood
  • Order the acrylic sheets

This will ensure we stay on track for upcoming milestones and allow us to refine the system before final integration.

Posted on

Team’s Status Report for 02/22/25

Risks

  • Running ML models on Raspberry Pi could cause performance limitations due to limited computing power. To address this, I will optimize models for efficiency and consider offloading computations to an external server.
  • Additionally, compatibility between ML frameworks and Home Assistant might pose challenges.I will validate API integrations using Postman to address this.

Changes

No major changes in implementation/design yet. After trying to deploy the ML model on RPi, if the RPi has limited storage/processing power, will deploy the ML forecasting on a computer.

Progress

  • Automation setup with Home Assistant.
  • Initial implementation of optimization models.
  • Model training and prediction framework.
  • CSV data storage and Docker integration.
  • Nordpool grid price retrieval setup.