Infra8
Case studyOur productEnergy

Solar Atlas: A Solar Dashboard on a Multi-Environment AWS Platform

How we built Solar Atlas, a live dashboard and peak-hour scheduler for hybrid solar inverters, on AWS with Terraform and pull-request deployments.

Product overview

Solar Atlas

Solar Atlas connects to a hybrid inverter managed in SmartESS or DessMonitor, shows the live state of the house, stores the history, and switches the inverter between modes on your peak-hour schedule.

Live energy flow between solar, battery, grid and home
Live energy flow: the battery is powering the home
  • Live energy flowSolar, battery, grid and home in one animated diagram, with stat tiles for charge, voltage, frequency and load.
  • History that staysEvery reading is sampled every 5 minutes and stored, so you can look back at any day.
  • Peak-hour schedulerSet your peak window once. Solar Atlas applies Peak or Off-Peak mode at each boundary, even with the dashboard closed.
  • Three inverter modesDefault, Peak hour and Off-peak hour. Apply one manually and it holds until the next boundary.

Works with hybrid inverters managed in SmartESS or DessMonitor. Independent of, and not affiliated with, those vendors. Screenshots show the real dashboard with sample readings.

Opens solar.infra8.co in a new tab

Continue to Solar Atlas
Partner With Us
Solar Atlas energy flow with the battery powering the home, beside the battery and home load readings
The real dashboard, shown with sample readings.
Type
Our own product
Built by
Infra8, in-house
Runs on
Amazon Web Services
Area
Energy monitoring and automation
Stack
The challenge

Why We Built It

Short answer

Solar Atlas is our own product: a live dashboard and peak-hour scheduler for hybrid solar inverters. It started as a tool for one home installation and became our reference AWS platform, with separate dev and production environments defined in Terraform, every change planned on a pull request, and an architecture that keeps running costs low.

Hybrid inverter apps show what is happening now, but keep little history and have no time-of-use automation. Making the most of cheaper off-peak power meant switching the inverter's mode by hand at every peak boundary, every day.

The engineering problem came next. The first version lived in a single production-only setup, so every change touched production and nothing could be rehearsed first.

What we built

What Solar Atlas Does

  • 01

    Live energy flow

    Solar, battery, grid and home in one animated diagram, with the readings that matter beside it.

  • 02

    History that stays

    A reading is stored every five minutes, so any past day can be charted.

  • 03

    Peak-hour scheduler

    Set the peak window once. Solar Atlas switches modes at each boundary, even with the dashboard closed.

  • 04

    Three inverter modes

    Default, Peak hour and Off-peak hour, applied on schedule or by hand.

  • 05

    Installable web app

    Works in any browser and installs to a phone's home screen like an app.

How it works

The System, Step by Step

A high-level view of how the pieces fit together.

How readings and schedules flow
  1. 01 · SourceInverter cloudSolar Atlas uses the inverter's existing cloud account to read data and change modes.
  2. 02 · ServiceBackend and schedulerA small containerised service on AWS serves live data and switches modes on time.
  3. 03 · CollectScheduled readingsA serverless function stores a reading every five minutes.
  4. 04 · StoreHistory databaseReadings are kept in a managed, pay-per-use database.
  5. 05 · ShowDashboardAn installable web app delivered through a global CDN.
How a change reaches production
  1. 01 · ProposePull requestEvery infrastructure change starts as a pull request, with the Terraform plan posted for review.
  2. 02 · CheckAutomated checksFormatting, linting, security and secret scans run before anything merges.
  3. 03 · ApplyDev on mergeMerging applies the change to the dev environment first.
  4. 04 · PromoteProduction on purposeProduction is updated in a separate, deliberate step.
Engineering decisions

The Choices That Shaped It

  1. 01

    One set of modules, separate environments

    Dev and production are built from the same Terraform modules, each with its own settings and state, so a change can be rehearsed before it reaches production.

  2. 02

    No stored cloud keys

    The pipeline signs in to AWS with short-lived credentials, so no long-lived access keys sit in CI.

  3. 03

    Cost-first architecture

    No load balancer, NAT gateway or Kubernetes where a small service does the job. Serverless functions handle collection and history.

  4. 04

    A private origin for the web app

    The dashboard's files sit in a private bucket that only the CDN can read.

Inside the product

Screens From the Product

Real product screens, shown with sample data.

  • Live energy flow between solar, battery, grid and home
    Live energy flow: the battery is powering the home
  • Battery, solar, grid and home load readings, each with its status
    Readings for battery, solar, grid and home
  • Peak and off-peak schedule, automatic switching, and the three inverter modes with Peak hour active
    The peak-hour schedule and the three inverter modes
What it shows

What This Work Demonstrates

  • A dashboard that keeps the history the inverter app doesn't, and switches modes on schedule.
  • A reference AWS platform: the environments, pipeline and reviews we bring to client cloud work.
  • Infrastructure that is rebuilt from code, and can be torn down when it isn't needed to control cost.

Screens show the real dashboard with sample readings. Solar Atlas is independent of, and not affiliated with, SmartESS, DessMonitor or Eybond.

Related service

Cloud/DevOps management

Your cloud, CI/CD and security run for you. Starts with a free audit.

Explore Service
Keep reading

More Case Studies

Get started

Tell Us Where Your Product Is. You'll Hear Back in 24 Hours on Business Days.

Partner With UsExplore Services
  • Fixed scope
  • You own everything
  • NDA on request