EA F1 × Lumi AI development announcement with gold and cyan motorsport technology imagery for demo research

Development started: EA F1 Auto × Lumi AI Auto — Build the system, follow the plan, measure the data

EA F1 × Lumi AI development announcement with gold and cyan motorsport technology imagery for demo research

Development has started — EA F1 Auto × Lumi AI Auto is a project to turn clear trading rules into a system and use data to examine how AI can help select when a new cycle may begin. The project is in development and will begin testing on Demo accounts. It is not publicly available, and there are no project test results to report yet.

A simple idea: EA follows the rules, Lumi filters the time window

We are preparing to compare two Demo systems with the same entry and exit rules and the same method of setting risk. Lumi AI Auto adds one decision gate: allow or skip START during the current window. Permission only opens that gate; it does not trigger an immediate START. Every signal must still satisfy the playbook.

TopicDemo A · EA F1 Auto baselineDemo B · Lumi AI Auto
Starting a cycleTake every signal that passes the playbook to establish a comparison baselineSTART only when AI allows it and the signal passes the playbook
Signal checkWait for the yellow candle to close; the button must not be blueUse the same conditions as Demo A
AI reviewNo AI gateAnalyse a chart snapshot every 15 minutes; choose “allow” or “skip”
During a cycleDo not START again while a cycle is runningNo repeated START; a new AI decision does not change an open cycle
Risk and exitsEnd the cycle at SL / TP / price breaking the orange line, as defined by the latest EASame entry, exit and risk-setting method; first-version AI does not change lot size, direction or exit rules
Data failureLog technical issues separatelyStale data or an unresponsive API: block all new STARTs and log the reason

Two layers before a new START

Layer 1 · Lumi AI (Demo B only)

Snapshot every 15 minutes → allow / skip

Layer 2 · EA rules and playbook (A and B)

Yellow candle closed + button not blue + playbook passed + no active cycle → START

Demo A begins at layer 2. Demo B must pass both layers. A skip, stale data or API failure blocks all new STARTs in that window. An open cycle keeps its original cycle-ending rules.

What do allow, skip and cycle end mean?

  • “Allow” does not mean start immediately: all EA and playbook conditions must first be met.
  • “Skip” means no new STARTs anywhere in that window, even if a signal appears. It does not change the rules or close an existing cycle.
  • Neither system repeats START during an active cycle. SL / TP / the orange-line break follow the latest EA definition frozen for that test period; we are not inventing an additional exit definition.

Memory lives in the database; development must remain traceable

We plan to store the image taken before each decision, indicator values, the decision and its reason, orders, and the net result of the entire cycle, including skipped signals. This database is the system’s memory for review and analysis. Sending images to an API does not mean the API permanently learns from them by itself.

Chart patterns will be analysed to propose and test additional filters. Data collection continues, but the active filters remain frozen within each test period. A new filter must be tested on a new data period before adoption. Versions will be recorded and reversible, so we can identify what changed and what effect it had.

Planned reports: measure both the benefits and the tradeoffs

  • Daily and weekly: net results after costs, Maximum Drawdown (Maximum DD, the largest decline from a portfolio peak), and cycle count.
  • Measure losses avoided and profits missed through skipping, so the review does not select only favourable outcomes.
  • Separate portfolio results from per-signal comparisons, and explain how skipped signals are compared.
  • Separate technical failures, such as stale data or API timeouts, from AI decisions to skip.

Status and limitations, stated clearly

This is development and preparation for Demo research. No launch date is being announced, and public access is not open. Good short-term results would not establish an edge. Demo results do not guarantee live-account results, and the system can lose money.

Read the foundations and follow progress

The EA F1 page and guide below explain the existing system. This announcement defines a separate research scope for EA F1 Auto × Lumi AI Auto. Example settings in the guide are not automatically the settings for this project’s tests.

EA F1 · EA F1 Semi-Auto Playbook

Follow progress on Paaoh. If you are interested, send “F1” through LINE @paaoh, the website’s existing contact channel, to register your interest and ask about development progress. This is not an invitation to start using a publicly released system.

Send “F1” via LINE @paaoh

Comments (

0

)

Discover more from Paaoh

Subscribe now to keep reading and get access to the full archive.

Continue reading