Energy Coordination for Manufacturing

Protecting margins through load-shedding, diesel cost reduction, and production optimisation

Grid ResilienceDiesel OptimisationProduction Scheduling

The Energy Problem

Industrial facilities in emerging markets already own the assets: grid connection, diesel generators, maybe solar and batteries. They just don't work together.

Load-shedding destroys production schedules. Diesel costs 3-5x grid electricity and crushes margins. Equipment sits idle waiting for power, then runs at peak cost when it shouldn't. Cold rooms lose temperature. Production misses deadlines.

Underscore coordinates your existing energy assets (grid, diesel, solar, batteries) around your production requirements. The result: reduced diesel costs, protected margins, and production reliability through grid failures.

PROD

Energy-Intensive Production

Coordinating grid, diesel, and solar across production equipment to protect margins

Bakeries & Milling

Proof Point

Model: Industrial bakery producing 1M+ loaves daily, reducing energy costs through smart grid/diesel coordination

Challenge

Ovens need preheating, which is energy-intensive. Grid schedule says 4pm-10pm but actually comes 6pm-8pm (maybe). Diesel costs 3-5x grid electricity. A facility running all-day shifts on diesel burns through fuel budgets rapidly. How do you plan production when you don't know when electricity will be available, without destroying margins?

How Underscore Helps

Edge controllers track energy requirements and efficiency at each load. The orchestrator routes work to equipment with cheapest available power at that moment. When grid power is available, production shifts to the grid window. When grid fails, controllers continue autonomously on diesel, queuing non-urgent work for later. The system learns your grid schedule and optimises production timing to minimise diesel use.

Cassava & Root Crop Processing

Proof Point

DRC produces 45M+ tonnes cassava annually (FAOSTAT), most processed inefficiently or lost to post-harvest spoilage

Challenge

Processing is energy-intensive. Seasonal input gluts create pressure to process quickly. Limited capacity during peak harvest. How do you maximise throughput when both supply and power fluctuate wildly?

How Underscore Helps

Processing zones expose their capacity, energy profiles, and scheduling constraints. The orchestrator optimises which inputs go where and when, factoring in power availability and cost. Pre-processing shifts to grid windows. Throughput is maximised; diesel use minimised.

COLD

Cold-Chain & Refrigeration

Managing refrigeration loads around power availability to maintain quality whilst controlling costs

Dairy & Milk Processing

Proof Point

Model: 500K-1M litres/day with guaranteed cold-chain integrity and reduced energy costs

Challenge

Refrigeration can consume 40% of facility energy costs. Temperature must be maintained at every step. Diesel backup costs multiply. Any break spoils the batch. How do you maintain quality whilst managing energy costs across distributed processing and logistics?

How Underscore Helps

Every stage (processor, storage, transport) is instrumented. Real-time temperature and energy consumption data flows to the orchestrator. The orchestrator optimises refrigeration timing around grid windows: pre-cooling when grid power is available, coasting through outages, minimising diesel use. If any zone deviates from temperature range, it's detected instantly and the batch is rerouted or salvaged.

Cold Storage Facilities

Proof Point

$4B+ annual loss from post-harvest and cold-chain gaps in Sub-Saharan Africa (FAO)

Challenge

Cold rooms are the largest energy load in many facilities. Running them on diesel during load-shedding destroys margins. But letting temperature rise destroys product.

How Underscore Helps

Thermal mass is an asset. The orchestrator pre-cools aggressively when grid power is available, building a thermal buffer. During outages, it manages which cold rooms get diesel backup based on product value and temperature sensitivity. Cold-chain becomes a managed, cost-optimised system rather than an emergency.

RES

Grid Resilience & Diesel Optimisation

Maintaining operations through load-shedding whilst minimising diesel consumption

Production Through Load-Shedding

Proof Point

Real: DRC cities experience 10-15 hours daily without grid power (World Bank). Underscore controllers maintain autonomous operation through grid failures

Challenge

Power is unpredictable. You can't just flip a switch when grid returns. Equipment needs startup sequences. How do you plan production when you don't know when electricity will be available?

How Underscore Helps

Controllers are autonomous. They execute workloads independently and maintain state locally. When grid is available, they use grid power. When grid fails, they continue on backup power for critical loads or queue non-urgent work. The orchestrator learns grid patterns and pre-positions the system. When connectivity returns, all data syncs. Production never stops.

Hybrid Energy Coordination

Proof Point

Facilities with grid, diesel, and solar often can't coordinate them effectively. Each system operates independently

Challenge

You have grid (unreliable, cheap when available), diesel (reliable, expensive), and maybe solar (intermittent, free). How do you make them work together rather than fighting each other?

How Underscore Helps

The orchestrator sees all energy sources as a unified system. It predicts grid availability, solar output, and production demand. It pre-charges batteries from grid or solar, shifts flexible loads to cheap power windows, and reserves diesel for critical loads during extended outages. Your accidental microgrid becomes a coordinated energy system.

Seasonal Grid Stress & Solar Optimisation

Proof Point

Sri Lanka hydro dropped to ~20% of output during the February 2025 dry spell (CEB). Industrial tariffs proposed to rise 11.57% for Q1 2026 (PUCSL)

Challenge

In hydro-dependent grids (Sri Lanka, East Africa, Southeast Asia), load shedding is seasonal rather than chronic. Dry seasons drop hydro output, tariffs surge, and factories fall back on diesel. Rooftop solar is installed but uncoordinated with backup generation and grid tariff windows. How do you minimise diesel runtime when the grid is unreliable for months at a time?

How Underscore Helps

Disaggregation identifies which loads are flexible from a single meter point. The orchestrator shifts flexible loads to solar hours, minimises diesel runtime during shedding windows, and pre-positions before tariff peaks. When the rains return and grid stabilises, the system adapts automatically without manual reconfiguration.

NET

Supply Chain Networks

Coordinating distributed production and distribution as a unified system

Informal Retail Networks

Proof Point

Model: Thousands of independent street vendors and kiosk operators coordinating as a distributed network

Challenge

Independent retailers operate in isolation. Each knows their neighbourhood but has no visibility into adjacent areas' inventory or demand. Producers have no way to see which areas will be overstocked or short. How do you coordinate thousands of independent actors without centralising control?

How Underscore Helps

Each retailer becomes a node, reporting stock and sales in real time via simple tools (SMS, app, or radio). They remain fully independent, setting prices, serving customers, making decisions. But the orchestrator sees the whole network: which areas are overstocked, which are running short, where demand is shifting. It suggests restocking patterns and coordinates production timing. Retailers gain visibility; the network gains coordination.

Multi-Site Production

Proof Point

Multi-city deployments share learning. Data from Kinshasa improves operations in Nairobi

Challenge

Multiple production sites, multiple energy situations, multiple local constraints. How do you optimise across sites rather than within each one?

How Underscore Helps

Each site is a node in a larger network. The orchestrator sees capacity, energy costs, and demand across all sites. It routes production to facilities with lowest energy costs at that moment. Patterns discovered at one site improve forecasting at others. Network effects compound.

VIS

Visibility & Optimisation

Making energy consumption and production flows visible and continuously optimisable

Real-Time Energy Visibility

Proof Point

Most facilities don't know their energy cost per unit of production, just the monthly bill

Challenge

Energy costs are buried in monthly invoices. You can't optimise what you can't see. Which equipment is inefficient? Which production runs cost more? Where is diesel being wasted?

How Underscore Helps

Every controller logs energy consumption against production output. The orchestrator aggregates this into real-time cost-per-unit visibility. You see which equipment, shifts, and products consume most energy. Inefficiencies become visible. Optimisation becomes possible.

Waste Prevention

Proof Point

$4B+ in post-harvest losses across Sub-Saharan Africa annually (FAO). Much is preventable with better coordination

Challenge

Producers over-make. Distributors over-stock. Product expires. Waste happens silently because nobody sees the whole system.

How Underscore Helps

Controllers track inventory and shelf-life in real time. The orchestrator sees the entire network and matches production to actual demand. Automated routines re-price or reroute product approaching expiry. Demand signals propagate upstream. Waste becomes visible and preventable.

Core Architecture

Every use case (production, cold-chain, resilience, supply networks) is solved by the same underlying architecture.

Defined Boundaries

You define logical boundaries around production stages, zones, or outcomes. Each zone exposes state (energy consumption, production output, status) through standardised interfaces. The same software coordinates baking, cold storage, or packaging.

Distributed Coordination

Edge controllers share predictions, negotiate priorities, and coordinate energy routing and production scheduling. There's no single point of failure. Decision-making is distributed across controllers, each running logic locally.

Edge Autonomy

Each controller executes locally and maintains state. During infrastructure failures (grid down, network down), controllers continue running in their zones. When connectivity returns, they sync and re-coordinate. One controller failing doesn't bring down the system. Production doesn't stop.

Network Learning

Every controller produces data. The orchestrator aggregates it. Patterns discovered in one deployment improve all others. The network learns continuously.

Integration Without Replacement

Controllers connect to existing equipment. They don't replace it. Your 50-year-old oven and your new production line both become visible and coordinatable. Deploy a controller, gain control.

Parameterised Deployment

The core architecture remains consistent. What changes are the parameters: local grid schedules, energy costs, equipment profiles. The system adapts to context without reinvention.

Make Your Facility Programmable

Whether you operate production facilities, manage cold storage, or coordinate distributed manufacturing, Underscore integrates with your existing operations. No replacing your equipment. Better energy coordination. Better margins.

Discuss Your SystemLearn the Architecture