The XPower project's learning environment
Agents learn. People find their bearings.
Energy, intelligence and space form one system. Keeping the entire system in one human mind is practically impossible.
That is why the academy is built primarily to train agents. It gives people a map and points of reference. Agents are responsible for mastering the body of knowledge, performing calculations and checking decisions.
People / orientation
Understand how the system works, choose your task and assess an agent's conclusions. People do not need to master the entire body of knowledge. A first pass takes approximately 20–30 minutes.
Start the orientation →Agent / main learning path
Master the subject, follow its prerequisites, reproduce the calculations and verify the sources. Agents perform the academy's main learning and calculation work. Results must allow independent verification.
Open the learning map →The human task is to find a place in the system, pose a specific question and assess the consequences of a decision. Agents take responsibility for mastering the subjects and performing the calculations.
The next-generation energy engineer: introductory part and competence map.
The fifth group: a shop, pumping station, building, transport or communications. What does the facility do for people, and what new role could it take on?
Six international case studies: what is confirmed, what stopped, and what questions remain.
Choose a facility, a useful result and constraints. Ask for a calculation, the source of every important number and the conditions for stopping.
What physically changes at the facility?
Who receives a useful result, and what do they pay for?
Which data can verify the promise?
The first-pass time is approximate. Professional work in any discipline requires deeper training.
Six pathways that turn knowledge into completed work. Books, calculation laboratories and projects with verifiable outcomes.
6 workshops · 48 chapters · 36 exercises · 3 languages
Open workshop →Twelve entry points into one system. Start with the overall picture, then move from the physical facility to control, economics and planetary scale.
A generalist connects physics, economics, law and computing to solve one task.
How things work and what can be changed.
02 · Physics and equipment
Refrigeration systems, useful heat, equipment service life and operating control.
Learning outcome: reproduce the heat balance and explain the limits of savings.
Open the textbook →03 · Behavior Management
Coordinated load changes across many facilities. The international term is Demand response.
Learning outcome: calculate a load-reduction event, including recovery and verification of the result.
Open the textbook →04 · Power, time, behavior
Peak costs and load balancing: from a kettle and a shop to a power system and a space station.
Learning outcome: check the physics of load shifting, storage power and energy capacity, costs and recovery.
Open the textbook →05 · Connection Architecture
How physical flows, data, rights and settlements connect distributed nodes.
Learning outcome: describe connections, constraints and actions when communication is lost.
Open the textbook →06 · Life Infrastructure
Food, water, temperature, air and light. New roles of familiar objects and a master plan for transition.
Learning outcome: prepare a node record and a plan to verify its new energy role.
Open the textbook →07 · Autonomy and vital functions
How to preserve life and a functioning environment. 72 hours without the external grid: functions, energy, authority and recovery.
Learning outcome: verify 72 hours of autonomy: functions, energy, power, logistics, authority and recovery.
Open the textbook →How a result becomes a product and an obligation.
08 · Data and controllability
Measurement, forecasts and available actions as products. Tensor: facility data across time, operating modes and constraints.
Learning outcome: describe the product, authority and verification of delivery without selling the same resource twice.
Open the textbook →09 · Rights and settlements
How a verified physical result becomes an obligation and a digital right.
Learning outcome: examine the debtor, basis for payment, measurement and risk of a specific product.
Open the textbook →10 · Composite Trading
An energy price linked to the price of the final product. Technology, intelligence and the allocation of verified results.
Learning outcome: build a pricing formula and show who bears the risk and receives the benefit.
Open the textbook →How to connect energy, computing and space.
11 · China · India · open networks
From Jeremy Rifkin's Energy Internet concept to Chinese infrastructure and India's open digital networks.
Learning outcome: compare a built system, a pilot, a protocol and a policy initiative.
Open the textbook →12 · Energy, communication, computation
A combined Earth–orbit balance. Solar energy, power transmission, computing and heat rejection.
Learning outcome: verify the energy balance, mass, communications and economics of an orbital scenario.
Open the textbook →Three connected studies of long-duration autonomy: the power system, a store at −30 °C, then community supply and food safety.
Arctic and Antarctic diesel, solar, wind, storage, heat and safe diesel-off operation.
Outcome. A testable polar microgrid architecture with balances and stop conditions.
Temperature, humidity, cold chain, fuel, failures and 30–60 days of operation.
Outcome. A store plan covering resources, deliveries and safe recovery.
A warm core, product relocation, shared storage, stock distribution and northern community practice.
Outcome. A food-hub record and a testable 60-day supply plan.
Status describes the result as of the source date. An operating system, pilot, experiment, plan and lost mission call for different conclusions.
The UHVDC line is in operation. The supplier's description states: 12 GW, 3 293 km and 1 100 kV.
Lesson. Study equipment, losses, line load and execution organization together. Digital layer relies on physical infrastructure
Which part of the result transfers to a network of small distributed facilities?
Hitachi Energy · project participant ↗Tata Power-DDL reported the launch of a pilot with 65 prosumers, 75 consumer sites and more than 2 MW of solar capacity.
Lesson. A digital transaction requires grid participation and settlement rules. The scale of a 2021 pilot cannot automatically be extrapolated to today's market.
What else is needed for the market to scale once data exchange works?
Tata Power-DDL · pilot operator ↗MAPLE transmitted energy to receivers in space; a transmission signal from orbit was detected on Earth.
Lesson. The experiment validates a specific mechanism. Industrial power levels, delivery costs and system service life need evidence of their own.
Which next experiment would bring this result closer to practical energy supply?
Caltech · team report ↗An architecture for computing in space was presented. The publication schedules the launch of two prototypes with Planet for early 2027.
Lesson. Solar power, computing, optical communications and heat rejection are designed together. A planned mission retains its status as a plan.
At what mass, launch price and communications availability do the economics work?
Google Research · authors of architecture ↗NASA links navigation error to inconsistency of English (inch-pound) and metric units in software data.
Lesson. Units and conventions between systems must be checked automatically. Energy and power data require the same discipline.
Which check would let your agent stop a similar error before execution?
NASA · mission description ↗According to the Tesla Science Center's history, the tower never became an operating system; the project stopped amid financial difficulties and doubts about its commercial viability.
Lesson. A major idea needs a measurable intermediate result and funding for the next stage. The project history helps us test our own promises.
What minimum result could be demonstrated before building the entire system?
Tesla Science Center · project history ↗University courses provide a foundation. Research programmes show the frontier. Open models let you test the reasoning yourself.
How to manage a complex system from requirements to testing and operation.
First step: Start with the life cycle, requirements and the distinction between verification and validation.
Mission operations, reliability, ECSS standards and the economics of space projects.
First step: Open E-Learning. For in-person programmes, check admission requirements and enrolment.
Engineering problem definition, flight and aerospace systems design.
First step: Complete the design overview and one assignment, then move on to current missions.
Microwave power transmission, beam control and space energy experiments.
First step: Record the power, distance and conditions of each test.
Electrical machines and systems. Lecture notes, textbook, exercises and solutions.
First step: Study the section on power and one worked calculation exercise.
Smart grids, from basic architecture to modern control technologies.
First step: Start with smart grid architecture; check certification requirements separately.
Solar energy, storage, electrical grids and the energy transition.
First step: Choose one technology and relate its parameters to a model of the facility.
An entry point to the Chinese research community working on the Energy Internet.
First step: Review the institute's research areas and compare them with the academy's Chinese case studies.
Electricity for computing, infrastructure constraints and AI applications in energy.
First step: Read the executive summary. Record the date and assumptions of each forecast.
Modelling power systems, markets, storage and investments.
First step: Reproduce the Quick Start and change one constraint. Save the model version.
A common language for digital interactions in energy: offers, orders and delivery.
First step: Read the README and a transaction example. Record the specification version.
Agent tools, task execution and evaluation of results.
First step: After the introduction, build an agent for one verifiable learning case.
The format and language are listed beside each link. Course materials, certificates and in-person participation may have different access requirements. The selection includes foundational courses from earlier years and current research materials.
The map sets the order of topics, prerequisites and expected outcome for each textbook. Start with the generalist, then study physics and control. Test autonomy in energy systems in disasters. Then move on to data, rights, trading and orbit.
Reading the materials alone does not establish readiness. Assessment requires a calculation and a new task with changed conditions.
Replace the facility and result with your own, then provide the text together with the academy's address.
Read the map at https://x5power.com/Academy/navigation/curriculum.json. Facility: a shop with a refrigeration system. Task: assess available flexibility without breaching temperature limits. Study the required textbooks and primary sources. Return: input data, calculations with units, constraints, load recovery, economics, sources and conditions for refusing execution. List missing data separately.
Verification protocol: 20 traps, 12 scenarios and 16 checks before delivering a result. Source, unit, boundary, stopping conditions and independent verification.
Learning calculations and decision checks use models. Passing criteria are provided in the learning map. Moving on to control real equipment requires separate engineering acceptance.