Future Considerations
Features explicitly deferred to maintain greenfield velocity. Revisit when triggering conditions are met.
API & Infrastructure
| Feature | Rationale for Deferral | Triggering Condition |
|---|---|---|
| API versioning strategy | Greenfield: no existing clients to support | Customer count > 100 OR external API consumers emerge |
| Rate limiting & quotas | No abuse patterns observed yet | Sustained API abuse OR customer tier differentiation needed |
| Webhook system | No customer requests for event notifications | 3+ customers request webhooks OR integration ecosystem emerges |
Data & Analytics
| Feature | Rationale for Deferral | Triggering Condition |
|---|---|---|
| Report generation engine | Elasticsearch ad-hoc queries sufficient | Customers request scheduled/automated reports |
| Historical data aggregation | Current time-series performance adequate | Query performance degrades OR storage costs exceed threshold |
| Advanced analytics (ML/AI) | Deterministic strategies prioritized | Proven ROI for ML-based optimization in pilot deployments |
Enterprise & Compliance
| Feature | Rationale for Deferral | Triggering Condition |
|---|---|---|
| SSO providers (SAML, OAuth) | Keycloak handles current needs | Enterprise customers require specific SSO integration |
| Advanced RBAC/permissions | Basic role model sufficient | Complex organizational hierarchies OR audit requirements |
| Compliance certifications | No regulatory requirements yet | Target market requires ISO 27001, SOC 2, or GDPR certification |
| API marketplace | No third-party integration ecosystem | 5+ third-party integrations OR customer demand for plugin system |
Gateway & Edge
| Feature | Rationale for Deferral | Triggering Condition |
|---|---|---|
| OTA firmware updates | Manual updates acceptable at current scale | Gateway count > 100 OR update frequency increases |
| Batch gateway provisioning | Low volume deployment manageable manually | Deployment velocity > 10 gateways/week |
| Gateway high availability mode | Single point of failure acceptable in beta | Production deployments OR customer SLA requirements |
| Edge analytics & ML | Cloud processing sufficient | Network latency issues OR offline operation requirements |
Protocol & Integration
| Feature | Rationale for Deferral | Triggering Condition |
|---|---|---|
| Modbus RTU auto-discovery | Modbus TCP covers majority use case | Significant RTU-only device population emerges |
| IEC 61850 MMS/GOOSE | High library cost (€10k-50k) | Utility/substation integration customer signed |
| DLMS/COSEM (IEC 62056) | Smart meter protocols fragmented | Standardized meter integration requirement emerges |
| BACnet/IP protocol | Building automation niche use case | Customer requests for HVAC/building integration |
| Custom protocol SDK | No customer-specific protocols identified | Customer with proprietary device protocol |
Review Triggers
Reassess deferred features when:
- Scale threshold: Customer count > 100, Gateway count > 100, API requests > 1M/day
- Customer demand: 3+ customers request same feature
- Operational pain: Manual process takes > 4 hours/week
- Cost efficiency: Deferred feature saves > €10k/year in operational costs
- Competitive pressure: Major competitor launches similar feature
Long-Term Goals
Vision features for future platform evolution. No immediate timeline - pursue when core platform is stable and market demand is validated.
New Asset Types
| Asset | Description | Use Case |
|---|---|---|
| Wind Turbine | Small/medium wind generation with variable output profiles | Microgrids, rural installations, hybrid solar+wind sites |
| Heat Pump / HVAC | Controllable heating/cooling loads with thermal inertia | Demand response, load shifting, building energy optimization |
| Thermal Storage | Hot water tanks, ice storage, phase-change materials | Store excess solar as heat, shift cooling loads to off-peak |
Considerations:
- Wind requires weather data integration (wind speed, direction forecasting)
- Heat pumps have COP curves dependent on ambient/source temperature
- Thermal storage has slow dynamics (hours vs seconds for batteries)
Asset State Machine
Introduce formal state management for all asset types:
| State | Description |
|---|---|
Initializing | Asset starting up, not yet operational |
Ready | Operational, awaiting commands |
Running | Actively executing strategy/setpoint |
Standby | Low-power mode, quick resume |
Fault | Error condition, requires intervention |
Maintenance | Scheduled downtime, excluded from dispatch |
Offline | Communication lost, last known state cached |
Benefits:
- Strategies can skip faulted/offline assets
- Maintenance windows exclude assets from optimization
- State transitions logged for diagnostics
Simulation Enhancements
| Feature | Description | Value |
|---|---|---|
| Seasonal Solar Variation | Latitude/longitude-aware generation curves accounting for day length and sun angle throughout the year | Accurate annual yield simulation - January vs July can differ 3-4× at mid-latitudes |
| Fleet / Workplace EV Profiles | Daytime workplace charging (9 AM - 5 PM), fleet depot patterns (overnight charging windows), multi-vehicle scheduling | Test commercial/fleet scenarios beyond residential commuter pattern |
| Spot / Dynamic Pricing | Simulated day-ahead and intraday price volatility with realistic patterns (morning ramp, evening peak, weekend dips) | Test arbitrage and price-response strategies without live market data |
| Grid Connection Limits | Configurable MaxImportW / MaxExportW on SimulatedGridMeter representing contracted capacity | Test peak shaving, export limiting, and connection constraint scenarios |
Additional simulation ideas:
- Weather-correlated load profiles (HVAC load scales with temperature delta)
- Grid frequency simulation for frequency response testing
- V2G (Vehicle-to-Grid) bidirectional EV charging
- Multi-day scenarios with persistent state across simulation runs