Smart Asset Operations

Capability · Smart Asset Operations · Owner-grade Validation

Vendors sell the uplift. We validate the math.

Microgrid controllers, AI/ML forecasting, and predictive maintenance are three technologies that get sold separately and only deliver value as an integrated stack. The marketing decks promise eight to fifteen percent savings. The production data, when it gets shared, often runs two to four. We work alongside owners and operators deploying smart-asset technology to validate vendor claims before contracts get signed, lock in data portability so the platform doesn't own your operating history, and translate the uplift into pro forma terms a lender's IE will accept.

Reality check

Every vendor in this space has a customer logo wall and a case study deck. Few have third-party-validated production data at scale. The discipline that matters is the gap between the deck and the meter, and how you close it before signing the contract.

vs vendor white papers

Marketing claims are not operational results.

The case study was the best site under the best conditions for the best customer. The fleet average is rarely shared. We benchmark uplift against your asset, your tariff, and your weather, not the vendor's reference deployment.

vs in-house data science teams

Energy is a domain, not a hire.

Generic ML talent without sector context produces models that don't ship. Tariff structures, derate behavior, asset metadata, and IE expectations are domain knowledge before they are technical problems.

vs black-box SaaS platforms

If you can't audit the model, you can't trust the savings.

Closed platforms own your data, your model, and your switching cost. The "savings" become unauditable by design and the IE will discount them on that basis alone.

vs perpetual pilots

A POC that never scales is not a strategy.

Every site has run a pilot. Few have a capital plan, an organizational owner, and KPI realization tracking to take it past one substation, one feeder, or one site.

Seven workstreams. From use case to production scale.

Each workstream applies to all three technologies in the stack. The artifacts are vendor-agnostic, lender-grade, and portable to a successor advisor or in-house team if you ever bring this work in-house.

Use case definition and value pool sizing

What problems are we actually solving? Demand charge reduction, peak shaving, energy arbitrage, capacity firming, resilience, PdM ROI. Each use case sized in dollars, with a defensible baseline, before any vendor gets contacted.

You get: use case memo, value pool model, KPI definitions tied to financial outcome.

Vendor landscape and selection

The vendor universe across utility-grade and behind-the-meter platforms. Production deployments verified, references called, financial health checked, conflict-of-interest screen against your sponsor and lender. Selection on operational fit, not LinkedIn presence.

You get: vendor landscape memo, short list with rationale, RFI/RFP package, scorecard with weighting.

Data architecture and portability

Data ownership, format, API access, historical export rights, model portability. The contract clauses that determine whether you can leave the platform without losing your operating history. The clauses most owners don't ask for and later wish they had.

You get: data architecture memo, portability clause library, data audit checklist, exit-readiness plan.

Model validation and benchmarking

Independent backtesting of forecasting accuracy, savings attribution methodology, PdM false positive and false negative rates. Vendor claims subjected to the same test the lender's IE would apply on day one of diligence.

You get: validation report, backtesting methodology, savings attribution memo, IE-grade benchmarking package.

Integration with O&M and asset management

How alerts get routed, how recommendations become work orders, who owns each step of the workflow, how outcomes get measured and fed back. The technology only delivers if the human process around it gets built.

You get: integration playbook, RACI matrix, alert routing architecture, escalation framework.

Financial model and pro forma integration

Translating uplift into pro forma adjustments. P50 and P90 of AI-driven savings, sensitivity analysis, lender position on uncertainty bounds. We make sure vendor optimism doesn't end up in the underwriting case unmarked.

You get: uplift integration memo, scenario sensitivity, lender's IE briefing on AI/ML assumptions.

Production deployment and scaling plan

POC to production. Phasing across sites, capital plan, organizational ownership, KPIs for tracking realization. The plan that turns a single substation pilot into a portfolio-scale capability with a budget line and an owner.

You get: scaling roadmap, capital plan, KPI realization tracker, organizational change brief.

Smart-asset deployment gates · Cross-cutting capability gates

SG1

Use case lock

Value pool sized in dollars, KPIs tied to financial outcome, baseline established.

SG2

Vendor selection

Platform selected, contract negotiated for portability, audit rights, and exit terms.

SG3

Pilot validated

Backtesting closed, savings attribution memo signed, IE comfort secured.

SG4

Production scale

Full deployment funded, organizational owner named, KPI realization tracked.

Three technologies. One stack. Integration is where the value lives.

Pillar 01

Microgrid controller

Real-time orchestration of distributed energy resources. Solar, BESS, generators, loads, grid interface, EV charging. Dispatches against price, reliability, and emissions targets simultaneously.

Used for: demand charge management, peak shaving, islanding, capacity firming, behind-the-meter optimization.

Pillar 02

AI/ML forecasting

Predicts load, generation, and prices on hourly to multi-day horizons. Feeds the controller's dispatch decisions and the asset's revenue projections. Accuracy is the product; everything else is packaging.

Used for: day-ahead bidding, peak prediction, capacity allocation, hedging, revenue assurance.

Pillar 03

Predictive maintenance

Sensor data and ML models predict equipment degradation and failures before they occur. Inverter trips, transformer thermal events, tracker actuator failures, BESS cell drift.

Used for: O&M cost reduction, availability uplift, warranty claim discipline, capital planning.

Pillar 04

Integration and orchestration

Ties the three pillars into one operational picture. Single source of truth for KPIs, unified alerting, consistent data taxonomy, lender pack from one platform rather than three.

Used for: avoiding vendor sprawl, lender reporting, exit readiness, unified asset performance management.

What sits on your desk when we're done.

A complete smart-asset deployment file. Vendor-agnostic, lender-grade, portable. Every artifact transferable to a successor advisor or to your in-house team, with no proprietary methodology behind a paywall.

● Use case memo and value pool model Each use case sized in dollars with defensible baseline and uncertainty bounds.
● Vendor scorecard and selection rationale Production deployments verified, references called, conflict screen documented.
● Data architecture and portability terms Ownership, format, API access, historical export, model portability, exit clauses.
● Model validation and backtesting report Forecast accuracy, savings attribution, PdM error rates. IE-grade methodology.
● Integration playbook and RACI Alert routing, work order generation, escalation, ownership of each workflow step.
● Pro forma integration memo P50 and P90 uplift translated to pro forma terms, sensitivity, lender's IE briefing.
● Production scaling roadmap Phasing, capital plan, organizational owner, KPI realization tracker.

Considering a controller, forecasting platform, or PdM deployment?

Initial scoping calls are complimentary. Engagements range from single-platform validation (one vendor, one site) through portfolio-wide smart-asset strategy. Vendor-independent, conflict cleared, fee from the client only.

Start a smart-asset scope →