GridWatch - Real-Time Energy Cost Monitor for Small Manufacturers

Small manufacturing shops run CNC machines, compressors, and HVAC 24/7 without ever knowing which machine is eating their electricity bill. GridWatch connects to any smart meter or IoT clamp sensor, shows per-machine energy cost in real time, and fires Slack alerts when a machine spikes past its budget. Built for shop owners, not energy consultants.

𝕏 Post Reddit HN

Difficulty

intermediate

What is it?

Electricity is the second-largest operating cost for small manufacturers after labor, yet most shops have zero visibility beyond a monthly utility bill. GridWatch ingests data from cheap IoT current clamp sensors (Shelly EM, ~$30) already popular in maker communities, normalizes it against real-time utility rate APIs, and displays per-machine hourly cost on a dashboard any shop foreman can read. The target market is 250,000 small US manufacturers with under 50 employees who cannot afford enterprise energy management systems costing $50k+ per year. This is buildable in 3 weeks using MQTT for sensor data ingestion, TimescaleDB on Supabase for time-series storage, and a Next.js dashboard. The hardware dependency is a risk but the Shelly EM ecosystem is mature and the API is public.

Why now?

Shelly EM sensors dropped to $30 in 2025 and have a fully public REST API, making sub-$100 per-machine energy monitoring viable for small shops for the first time without custom hardware.

▸Real-time per-machine energy cost in dollars per hour from Shelly EM sensor data
▸Slack and email alert when any machine exceeds a user-defined daily cost threshold
▸Monthly savings report comparing actual vs. baseline energy spend
▸Machine schedule optimizer suggesting off-peak shift windows based on utility rate tiers

Target Audience

Small manufacturing shop owners with 5-50 employees, approximately 250,000 shops in the US spending $2k-$15k/month on electricity.

Example Use Case

Marco, owner of a 12-person metal fab shop, installs 4 Shelly EM clamps in an afternoon, sees his laser cutter is running idle at night costing $340/month, sets an auto-off schedule, and recoups the $79/month subscription in week one.

User Stories

▸As a machine shop owner, I want to see each machine's hourly electricity cost, so that I know which equipment to schedule off-peak.
▸As a shop foreman, I want a Slack alert when a machine runs past closing time, so that I can prevent overnight idle waste.
▸As a shop owner, I want a monthly savings report, so that I can justify the subscription to my accountant.

Acceptance Criteria

Sensor Ingest: done when live wattage data appears on dashboard within 10 seconds of sensor event. Cost Calculation: done when dashboard shows dollar-per-hour cost matching manual utility rate calculation within 5%. Alert: done when Slack message fires within 60 seconds of threshold breach. Monthly Report: done when PDF report generates with correct baseline comparison.

Is it worth building?

$79/month per shop x 50 shops = $3,950 MRR at month 3. $79/month x 200 shops = $15,800 MRR at month 9. Math assumes 3% conversion from cold LinkedIn outreach to local manufacturing groups.

Unit Economics

CAC: $150 via local in-person sales visits. LTV: $1,896 (24 months at $79/month). Payback: 2 months. Gross margin: 88%.

Business Model

SaaS subscription

Monetization Path

14-day free trial with one sensor. Paid tiers: $79/month for up to 10 sensors, $149/month for up to 30 sensors.

Revenue Timeline

First dollar: week 3 via beta shop conversion. $1k MRR: month 2. $5k MRR: month 6.

Estimated Monthly Cost

Supabase Pro with TimescaleDB: $50, Vercel: $20, OpenEI API: free, MQTT broker (HiveMQ free tier): $0, Resend: $20, Stripe fees: $20. Total: ~$110/month.

Profit Potential

Strong lifestyle business at $8k-$20k MRR. Acquisition target for energy SaaS or industrial IoT platforms.

Scalability

Medium — add predictive maintenance alerts, carbon reporting, multi-site dashboards for regional shop chains.

Success Metrics

Week 2: 5 beta shops with sensors live. Month 1: 3 paying customers. Month 3: $3k MRR.

Launch & Validation Plan

Visit 5 local machine shops, offer free sensor installation, validate they can read the dashboard without help, and ask for $79 before month-end.

Customer Acquisition Strategy

First customer: physically visit 10 small machine shops within 30 miles and offer to install one free Shelly EM sensor. Ongoing: LinkedIn ads targeting manufacturing business owners, local chamber of commerce events, partnerships with industrial electricians as referral partners.

What's the competition?

Competition Level

Low

What's the roadmap?

Feature Roadmap

V1 (launch): sensor ingestion, per-machine cost dashboard, Slack alerts, Stripe billing. V2 (month 2-3): monthly savings report, schedule optimizer, multi-sensor grouping. V3 (month 4+): predictive maintenance alerts, carbon report export, multi-site view.

Milestone Plan

Phase 1 (Week 1-2): MQTT ingest, TimescaleDB schema, dashboard UI — done when 3 sensors stream live data. Phase 2 (Week 3-4): alerts, Stripe billing, savings report — done when one beta shop pays $79. Phase 3 (Month 2): 5 paying shops, schedule optimizer shipped.

How do you build it?

Tech Stack

Next.js, MQTT.js, Supabase with TimescaleDB extension, Shelly EM API, OpenEI utility rate API, Stripe, Resend — build with Cursor for data pipeline, v0 for dashboard UI.

Suggested Frameworks

MQTT.js for sensor ingestion, TimescaleDB for time-series queries, Recharts for dashboard visualization

Time to Ship

3 weeks

Required Skills

MQTT protocol, TimescaleDB queries, Next.js, basic IoT sensor setup.

Resources

Shelly EM API docs, MQTT.js npm package, Supabase TimescaleDB guide, OpenEI rate API docs.

MVP Scope

pages/dashboard.tsx, pages/machines.tsx, pages/alerts.tsx, api/ingest.ts (MQTT bridge), api/machines.ts, api/alerts.ts, lib/timescale.ts, lib/rateApi.ts, supabase/schema.sql, vercel.json.

Core User Journey

Install Shelly EM clamp -> enter sensor ID in GridWatch -> see real-time machine cost in 5 minutes -> receive first overage alert -> upgrade to paid.

Architecture Pattern

Shelly EM sensor -> MQTT broker -> Next.js ingest API -> TimescaleDB -> dashboard queries -> Slack webhook alert if threshold exceeded.

Data Model

Shop has many Machines. Machine has many EnergyReadings (time-series). EnergyReading has wattage, cost, and timestamp. Shop has many AlertRules. AlertRule triggers Notification.

Integration Points

Shelly EM API for sensor data, MQTT.js for real-time ingestion, OpenEI for utility rate data, Stripe for subscriptions, Slack API for alerts, Resend for weekly digest emails.

V1 Scope Boundaries

V1 excludes: mobile app, predictive maintenance ML, multi-site management, carbon offset reporting, non-Shelly sensor brands.

Success Definition

A shop owner the founder has never met installs sensors, reads the dashboard, and pays the monthly invoice without any founder support call.

Challenges

Hardware installation is a sales blocker — shop owners want software but hate self-installing sensors. Offer a paid white-glove setup service ($199 flat) to close first 10 customers.

Avoid These Pitfalls

Do not build multi-site before getting 5 single-site customers. Hardware dependency means distribution is physical not viral — budget time for local sales. Finding first 10 paying customers requires in-person visits, not just ProductHunt.

Security Requirements

Supabase Auth with Google OAuth. RLS on all shop tables by shop_id. MQTT broker requires device token per sensor. Rate limit: 1,000 ingest events/minute per shop. GDPR: energy data deletion on account close.

Infrastructure Plan

Vercel for Next.js, Supabase Pro for TimescaleDB, HiveMQ free MQTT broker, GitHub Actions for CI, Sentry for errors. Total infra: ~$110/month.

Performance Targets

50 DAU at launch, 10,000 sensor events/day. Dashboard query under 300ms with TimescaleDB continuous aggregates. Page load under 2s. No Redis needed at launch.

Go-Live Checklist

☐Security audit complete
☐Stripe subscription tested end-to-end
☐Sentry live
☐Dashboard load tested with 30 days of mock data
☐Custom domain with SSL
☐Privacy policy published
☐3 beta shops signed off
☐Rollback plan documented
☐LinkedIn launch post drafted.

How to build it, step by step

1. Scaffold Next.js app and install MQTT.js, Supabase client, Recharts. 2. Enable TimescaleDB extension on Supabase and create energy_readings hypertable. 3. Write MQTT ingest API route that writes sensor payloads to TimescaleDB. 4. Build machine dashboard page with Recharts real-time line chart. 5. Write OpenEI rate lookup to convert wattage to dollar cost per hour. 6. Build alert rules UI and Slack webhook trigger. 7. Add Stripe subscription checkout for paid tiers. 8. Build monthly savings report page with baseline vs. actual chart. 9. Write machine setup wizard for entering sensor IDs. 10. Deploy to Vercel and test with one physical Shelly EM sensor.

Generated

April 14, 2026

Model

claude-sonnet-4-6

← Previous

LeaseParse - Plain-English Rental Lease Summarizer for Renters

SprintCast - Async Standup Bot That Actually Surfaces Blockers

Disclaimer: Ideas on this site are AI-generated and may contain inaccuracies. Revenue estimates, market demand figures, and financial projections are illustrative assumptions only — not financial advice. Do your own research before making any business or investment decisions. Technology availability, pricing, and market conditions change rapidly; always verify details independently.