Bali Solar Energy - Smart Energy Meter

A custom energy measurement system that shows how a property actually uses electricity through the day, so solar and battery systems can be sized from real data instead of estimates.

---

Project snapshot

 

---

The situation

A solar installation should match how a property actually uses electricity. That sounds obvious, but getting there is not straightforward.

A PLN bill tells you a property's total monthly consumption. A regular meter gives you a current reading. Neither tells you the thing that actually matters for solar: when the electricity is used across the day.

That gap has real consequences. Two properties can have almost identical monthly bills and completely different daily patterns. One uses most of its power during the day, when solar is generating. Another draws heavily in the evening and overnight, when it is not. The right solar and battery configuration for those two properties is different, even though their bills look the same.

This is the second system we built for Bali Solar Energy. Their AI sales system already turns a cold enquiry into a preliminary proposal built from four quick inputs, and offers the customer a free audit as the next step. This smart meter is that audit. It replaces the estimate with real measured data, so the final system is sized from how the property genuinely uses electricity rather than from a bill.

 

The challenge

The system had to collect real consumption data without turning a sales visit into an electrical engineering project. A potential customer should not have to modify their building's wiring just to be measured.

So the measurement had to be non-invasive, work on both common electrical setups (single-phase and three-phase), install without switching off the building's power, be simple enough for a trained employee to fit, connect to the cloud, run for several weeks, and not lock the company into one meter manufacturer.

That last point mattered most, and it shaped the whole approach.

 

Decisions before code

Two decisions defined the product before any development.

Do not touch the building's wiring. The meters use clip-on sensors, known as CT clamps, that read the electrical flow around the existing wires. Nothing is disconnected, the building keeps running, and a trained employee can fit it by following a short, documented procedure. That documentation can ship with the meter, so an installation does not always need a specialist on site. The result is a measurement system that is genuinely easy to deploy across many different properties.

Measure long enough to see the truth. A few minutes of data is not enough to understand the property's normal consumption pattern. Each property is measured for two to four weeks, which is long enough to reveal the real consumption pattern across different days and routines, rather than a single snapshot.

 

How the system works

The system runs in four steps, and the customer sees none of the complexity.

Install

The clip-on sensors go on without interrupting the power supply, on either single-phase or three-phase networks.

Connect

The meter sends its readings to the cloud over Wi-Fi. Where there is no Wi-Fi, it uses a 4G connection instead, and if both are unavailable it stores the readings locally to be collected after the measurement period. The system has multiple connectivity options, so a lack of Wi-Fi does not necessarily prevent measurement.

Measure

The meter records consumption continuously for the measurement window, building a clear picture of how the property actually uses electricity through the day.

Analyse

The readings flow into our own backend, which stores them, processes them, and runs the calculations for the solar assessment. The output is not a stream of raw numbers, it is usable information for designing the system.

 

From estimate to evidence

This is where the measurement becomes commercially valuable. It moves the sales and engineering conversation from:

"Your bill is about this much, so we think you need this system."

to:

"We measured how your property actually uses electricity, and we can size the system around that usage."

From real consumption data, the team can calculate the required solar capacity and use the daily pattern to shape the battery configuration. The aim is not to sell more panels. It is to match actual consumption, solar generation, and battery capacity to the property, which helps avoid both oversized and undersized systems.

 

Why the measurement is free

The measurement is provided free to customers planning to install their solar system with Bali Solar Energy, and the logic is simple: the better the data before the system is designed, the better the proposal.

Rather than charging separately for an energy assessment before the customer has even decided on solar, the measurement becomes part of the sales and engineering process. It removes a barrier for the customer and gives the company a real understanding of the property before making its recommendation.

 

The build

Rather than build measurement hardware from scratch, we assembled the system from commercial meters plus a custom backend and calculation engine. That keeps the hardware replaceable and avoids tying the whole platform to one supplier.

The measurement data does not stay in an engineering dashboard. The backend uses it to prepare the estimate and deliver a Bali Solar Energy-branded proposal, so the customer receives something they can act on rather than raw energy readings.

 

Release and validation

The system is currently in testing and validation. Three different commercial meter models are being evaluated against the things that will decide the final deployment: measurement accuracy, network connectivity, cloud API reliability, compatibility with different electrical installations, and installation simplicity.

The goal is not to declare the first working configuration finished. It is to find the hardware and workflow that can be deployed reliably across Bali Solar Energy's future customers. Once the preferred configuration is settled, the goal is to make the measurement infrastructure a repeatable part of relevant solar projects.

 

Outcome

The immediate outcome is a working IoT measurement system that can collect real consumption data from a property for two to four weeks without interrupting its power.

The measurement system in place. Photo from testing. No customer-identifying data shown.

It gives Bali Solar Energy a new way to size systems, from the property, to real consumption data, to analysis, to solar capacity and battery sizing, to a proposal.

Together with the sales system, this closes one connected funnel: an enquiry becomes a preliminary proposal, the free measurement turns that estimate into evidence, and the evidence sizes the final system. It also creates a lasting technical foundation. The hardware is replaceable, the data comes in through cloud APIs, and the calculations and proposal generation run in Bali Solar Energy's own backend. The company is not simply buying a smart meter. It is building its own measurement and solar-sizing infrastructure.

The system is still in testing, so we are not publishing accuracy or commercial performance numbers yet.

 

Why this is a techlead.fit case study

The starting point was never an IoT device. It was a business problem: how do you make a better solar recommendation when a monthly bill does not tell you enough about how a property actually consumes energy?

The answer required hardware, connectivity, cloud APIs, data storage, calculations, and a customer-facing proposal to work together as one system. We kept the installation simple, the hardware replaceable, and the data inside the client's own backend, so the value stays with their business rather than a supplier's.

Measure the real usage. Understand the pattern. Size the system from evidence. Keep the installation simple. Build the data into the business.

---

Want better technical decisions before you build?

This is the kind of project where the architecture has to be right before development starts. If you are considering an IoT product, connecting hardware to software, or building a system around real-world data, we can help define the architecture, validate the approach, and build the first working version.

Relevant services: Technical Audit · CTO Sprint · MVP Build