عربي
AI workflows · Illustrative case 01

SAGE: a smart agent for grid reinforcement.

A physics-aware AI workflow around the tools planners already use, DIgSILENT PowerFactory and Python. It decides what to study, reads the results, screens compliance and drafts the report. It doesn't replace the solver; it takes away the week of manual work around it.

Illustrative case. It shows what a workflow can do; it is not a client project.

Book a workflow demo Try it yourself
No.
01
Type
Multi-agent workflow
Tools
PowerFactory, Python
Agents
A coordinator and five specialists
Sign-off
An engineer at every gate
Built with
AI workflows
Status
Illustrative
The week

Most of a study week goes on the work around the solver.

Running PowerFactory takes minutes. Setting up the cases, sorting the results, checking compliance and writing the report takes days. SAGE hands that work to agents and keeps the engineer on the decisions.

40 h of engineer time on one reinforcement study

  1. Build and run the casesSimulation agent12 h
  2. Extract and sort the resultsData agent8 h
  3. Rank the reinforcement optionsDecision agent6 h
  4. Check N-1 and N-2 complianceCompliance agent5 h
  5. Draft the reportReporting agent9 h
  • Engineer time
  • Handed to an agent

Illustrative week for one reinforcement study. Each square is one hour.

See it run

One request, five agents, one engineer.

Watch a reinforcement study run from request to report, or pick an agent to see what it takes in and hands on.

Multi-agent run · interactiveIdle
COORDINATORAgentic coreAGENT 01SimulationAGENT 02DataAGENT 03DecisionAGENT 04ComplianceAGENT 05Reporting
The coordinator

Agentic core

Plans the study, splits it into tasks, hands them to the agents in the right order and sends every output to the engineer before anything is trusted. It organises the work; the engineer signs it off.

Takes in

  • Engineer's request
  • Study objective
  • Available agents

Hands on

  • Task plan
  • Agent calls
  • Run ready for review
The coordinator delegates · agents do the work · an engineer reviews every gate · every step is logged

Illustrative run. The figures are examples.

Try it

A data centre connects. You fix the network.

A 250 MW data centre asks to connect at substation F, and the N-1 study finds overloads. Pick reinforcements, run the study, and try to clear every overload for as little as you can.

N-1 study · 132 kV network–
Reinforcement options
–
N-1

Worst loading on each circuit

Not compliant
GSPGrid supply pointA80 MWB60 MWC70 MWDWind 120 MWFData centre · 250 MW
Worst N-1 loading need ≤ 100%
–
Overloaded circuits need 0
–
Cost
–
Cases

Each single outage

    Illustrative DC load flow on a six-substation network. A real SAGE run uses your PowerFactory model and your planning criteria.
    What it takes off your desk

    The work around the solver, automated.

    1. Automates

      Study set-up and runs

      Builds the scenarios and runs the batch, with no manual model wrangling.

    2. Automates

      Results triage

      Finds, ranks and explains the violations that actually matter.

    3. Automates

      Compliance and reporting

      Screens results against your criteria and drafts the report.

    4. You stay in control

      Engineer in the loop

      Your engineers review and approve every step before anything is trusted.

      You stay in control

    This is one example of what our AI workflows can do. Tell us the manual process that slows your team down, and we build the workflow that removes it.

    Next step

    Have a manual process that slows your team down?

    Book a workflow demo [email protected]
    Next illustrative case · 02 Stability Potential: stability services from below 132 kV