Six Sigma and Project Management · Grades 5 to 12
On the date, and right. Project management and Six Sigma taught as one subject, on the industry where a missed window costs twenty six months and a missed defect costs the mission. Personally, every class, one student at a time.
A process breathing under two fixed limits. Every number is computed from the curve.
01 The job nobody is taught
Count them. The science fair. The group history project with four people and one person doing it. The model rocket that had to fly on Saturday, the coursework due in eight weeks, the robotics season, the personal project that decides a grade at fifteen. A child between ten and eighteen is handed more projects than most adults are handed in a career, and is taught nothing whatever about how to run one.
So each is run on instinct. Started late, because nobody knows how long it will take. Grown quietly, because nobody wrote down what finished looks like. Finished the night before by the one person who cared, judged by feel, because nobody measured anything, and then blamed on the child, who is told to be more organised and more careful, as though those were character traits rather than techniques with names.
They are techniques with names. Define, plan, risk, people, measure, analyse, improve, flow, close. They were written down by the people who built the Polaris missile and the Apollo programme, and by the people who taught factories to make fewer mistakes, they fit on two pages, and a twelve year old can use every one of them on a launch day. The space industry uses them at their most unforgiving, because a window set by the planets does not move and the hardware gets one attempt, which is why it is the best possible place to learn them.
A teenager will happily learn critical paths, standard deviations and control limits without complaining once, because they are not doing statistics. They are working out whether the rocket is ready, and why it keeps going wrong. That is the most efficient teaching material I have found, and I have taught orbital mechanics to eleven year olds.
The students who come to me for this are rarely disorganised or careless. They are usually capable, overloaded, and tired of being the one who finishes everything at two in the morning and is then asked why it was not better.
Organised and careful are not personalities. They are nine techniques with names, and a ten year old can learn all of them.
Abhishek Dabas · MSc Astronautics and Space Engineering · BEng Aerospace Engineering
02 Where it all points
On the evening of 27 January 1986, engineers at the company that built the Space Shuttle's boosters argued by telephone with the managers who would decide whether it flew in the morning. The launch had already slipped. The forecast was for the coldest launch there had ever been, and the engineers believed the cold would stiffen the rubber seals in the boosters. One manager's question that night made it into the record: when did the engineers want him to launch, next April?
That is the schedule half of this programme, in one sentence. The quality half is what the engineers had in their hands: data from twenty four previous flights, and thirteen charts. Most of the thirteen did not put temperature on an axis at all. The one chart that would have made it obvious, every previous flight plotted against the temperature it flew in, the warm ones with no damage sitting beside the cold ones with damage, was never drawn that night. Whether it could have been drawn with the data actually in the room is still argued about, and a Stage Three student reads both sides of that argument.
That is the whole of this programme in one evening: a date arguing with a chart, and the chart not drawn well enough to win. The people were competent. The data existed. What was missing was two habits. A plan with enough room in it that a cold morning could wait a day. And the reflex of asking, before the meeting rather than after it, what picture would settle the question. This programme is those two habits, taught until they are reflexes, and it is the reason this starts in Grade 5 rather than in a first job.
03 The nine disciplines
A Gantt chart is what most people mean by project management, and statistics is what most people mean by Six Sigma. They are two of nine here. Four of the nine are how a mission reaches its window. Four are how the thing it delivers is right. The ninth is how you keep both.
What exactly, for whom, and what done means. Problem statements without a name in them, requirements that can be tested, the work broken down until each piece fits in a week. The Mars probe lost in 1999 because one team worked in pounds and another in newtons is the first story told here.
The order things must happen in, and how long. Networks, the critical path that decides the date, float, estimates as ranges rather than numbers, budgets and contingency. PERT was invented to get Polaris to sea; the arithmetic of why missions miss windows is on the page below.
What could go wrong, ranked. Risk registers and their engineering twin, the failure mode and effects analysis that every spacecraft carries; probability and impact, the four responses, and the pre-mortem. Apollo 13 came home on a contingency somebody had worked out years earlier.
Who cares, who decides, who does. Sponsors, teams, the go or no go poll where everyone answers out loud in turn, and the status report that says a date is in danger early enough for someone to do something about it.
Numbers you can trust. Operational definitions, sampling, whether the measuring itself is the problem, and a baseline in defects per million before anyone is allowed an opinion. Hubble's mirror was ground perfectly to the wrong shape because the instrument measuring it was assembled wrong, and two cheaper tests that disagreed were dismissed.
Why, proved. Pareto charts, fishbones, stratified data, scatter plots and hypothesis tests, and the variance between the plan and what happened. The difference between a cause and a coincidence, read the way an accident investigator reads a flight.
Change one thing and know it worked. Pilots, designed experiments, mistake proofing, and a before and after measured the same way both times. A rocket that learned to land did it by measuring every attempt and changing what the data said, not by trying harder.
The speed of work. Lean's eight wastes, value stream maps, takt and pull; agile's sprints, boards and retrospectives. Two vocabularies for one idea, work that moves instead of waiting, and the reason a launch pad can turn around in days rather than months.
Keep the gain, then hand it over. Control charts with limits drawn from the data, standard work, the checklist astronauts still use, formal acceptance, the post flight review written the week it ends, and the portfolio that turns eight years of projects into evidence.
Every term ends with a real project, delivered and measured, because that is the only form in which any of it is ever used. A plan, a risk register, a baseline, a proven cause, a fix that held, and a review presented to the person it was for.
04 The eight year plan
Three stages, eight years, nine disciplines in every one of them. Choose a stage and the whole map redraws itself at that depth. The full plan for each stage is a four page document, available on request.
The first stage is about finishing, and about noticing. A student learns that a project is anything with an end you can describe and a time it has to meet, that the end has to be described before the work starts, and that no two launches of the same rocket ever land in the same place. They run small real projects with a countdown: a model rocket launch day, an observing night, a science fair entry, planned on one page, flown, measured and reviewed. Every one of the nine disciplines appears here in a form a ten year old can hold, so that the words critical path and standard deviation stop being intimidating and start being interesting.
The middle stage converts finishing into method. Arithmetic, percentages, probability and a first standard deviation stop being schoolwork and become instruments: a student calculates a critical path rather than guessing it, plots a control chart rather than describing one, and learns that an estimate without a range is a wish and that a difference is not real until the variation says it is. This is where a student runs a real campaign, a rocketry season, a CanSat build, an observing programme, with a real schedule, real data and a real status report, and finds out that professional documents are written to be used under pressure rather than read for pleasure.
The final stage is taught at the level of the entry professional certifications and the Green Belt body of knowledge, and maps cleanly onto the mathematics and statistics a student is doing anyway for advanced placement, the international baccalaureate or A level. The purpose is a delivered mission: by graduation a student can take a sponsor, a team, a budget, a window and a dataset, produce a plan that survives contact with reality, prove with data that what was delivered works, report honestly when either is in danger, and explain what would have to change for the date or the conclusion to reverse. That is what the certifications test, and it is what the industry requires.
05 How it is actually taught
Two real lessons, compressed. They are the two I use most often, because each takes about ninety seconds to convince a thirteen year old that arithmetic, not effort, decides the answer. One is the schedule half of this programme. The other is the quality half.
Three teams build one spacecraft: the bus, the instrument and the flight software. Each makes a fair estimate, the kind where they are as likely to finish early as late. Nobody lies, nobody slacks, and each part is, on its own, a coin toss to be ready for the window. Ask a class what the chance is that the spacecraft is ready, and every hand says fifty percent.
It is one in eight. All three have to be ready, and the chances multiply. Add the launcher and the ground station and it is one in thirty two. Now give each team a chain of ten tasks, each estimated a little optimistically, the way everyone estimates, and the chance of the window falls through the floor, and the next window is twenty six months away. Nothing about the people changed. A plan is a product of probabilities, and products of numbers less than one only go one way.
Three sliders and one window. Everything on the right is derived from what you set on the left. Nothing here is decorative.
Move a slider.
Each task is modelled as five days give or take one and a half, which is roughly how real task durations spread. The plan promises the sum of the estimates; the picture rolls the dice every couple of seconds so you can watch the odds rather than take them on trust. In a real programme the ranges come from history, not from a model, and a student learns why that sentence is not a formality.
A rocket engine's injector plate has hundreds of drilled holes, and every one of them has to be a millimetre across, within a hundredth. Drill a thousand and measure them all: the average is one millimetre, exactly. The drill is, by the only number most people look at, perfect. And about two holes in every hundred are outside the limit, and a hole outside the limit is a streak of badly mixed propellant against the chamber wall.
Nothing is wrong with the average. What is wrong is the spread. Every process has a natural width, the distance its output wanders when nothing unusual is happening, and the drawing's limits do not care about your average, only about how much of that width falls outside them. The engineer's shorthand for this is capability: how many of your process widths fit inside the tolerance.
Three sliders and one tolerance. Everything on the right is derived from what you set on the left. Nothing here is decorative.
Move a slider.
A normal model of a drilling process, which is the model a student builds in the lesson and the one the published capability tables assume. The famous three point four per million belongs to a six sigma process after a 1.5 sigma allowance for drift; a Stage Three student learns why that allowance is argued about, and the readouts here do not use it.
06 What comes out of it
Not a worksheet and not a quiz score. Something with your child's reasoning written down in it, which is the only artefact that survives contact with a university admissions officer or an interview panel.
A real project: what done means, the work broken down, the chain that decides the date, the estimate with its range stated, and the current state measured properly, all before any work starts.
What could go wrong, ranked and owned; the weekly report that says where the project actually is, in numbers; and a Pareto, a fishbone and a test that confirm the cause with a number or abandon it in writing, which is the harder skill.
A change piloted and measured before and after, a control chart that shows it held after the student stopped watching, formal acceptance, and the post flight review written the week it ends.
A real project with a real sponsor in the space, aviation or astronomy world, a team, a budget, a window and a dataset. Planned as a professional plans it, tracked with earned value, its improvement proved with data, and reviewed at the end by someone who is not the teacher. Students who do this well never again call a deadline unrealistic without saying by how much and why, and never again describe a failure as carelessness.
07 The timing
Between fifteen and seventeen almost every serious curriculum hands a student a large independent project and very little training in how to run it or how to judge whether it worked. The Personal Project in the IB's middle years. The Extended Essay. Coursework for GCSE and A level. AP Research. The Duke of Edinburgh Award, a rocketry or CanSat season. These are graded, they matter to admissions, and they are the first time a student's ability to plan and to measure is tested rather than assumed.
In the same two years statistics turns up in school as a list of formulae to be applied to numbers that mean nothing to anyone. Standard deviation is taught as an algorithm, hypothesis testing as a ritual, and most students conclude, reasonably, that it is a subject for people who like ritual. A student who has plotted fifty control charts of their own rockets before meeting the standard deviation in a classroom treats it as an old friend, and that student chooses AP Statistics, or the IB's applications course, or A level further mathematics, rather than fleeing it.
There is a second effect, quieter and more valuable. A student who has spent three years saying early and precisely that a date is in danger, and proving with data that a fix worked, writes differently, argues differently and interviews differently. Selection panels for engineering courses and for graduate schemes are both looking for exactly that, and both find it very hard to teach.
By the time the big project arrives, a student either has a method or has a personality. Only one of those can be taught.
08 Where this leads
Project manager is the obvious one, and quality engineer the one most families have never heard of. Both were invented by the missile and moon programmes, and every industry now runs on them. These are the seats in the world this programme is taught on, and every one of them is opened by the same nine disciplines.
The part of a space programme whose job is to say no. Quality, reliability and safety on hardware that gets one attempt, at agencies, primes and launch companies, where the story at the top of this page is taught to every new engineer.
The people who run a spacecraft from proposal to launch, or a new aircraft from drawing to certification: systems engineering, schedule and budget, held in one head for a decade, against a window set by the planets.
Requirements, interfaces, review gates, and predicting when things will fail before they do. Statistics, physics and very careful record keeping, and the reason an engine is fired a hundred times on a test stand before it carries anyone.
Countdowns, checklists, go or no go polls and the flow of a launch campaign. Lean and human factors under a clock that does not stop, and the profession that turned a launch pad from a monthly event into a weekly one.
Some of the largest Lean and Six Sigma deployments on Earth build airliners, and an airline's on time performance is a control chart with passengers on it. Every discipline on this page, at the scale of a thousand aircraft.
Reading a system backwards from its worst day. Engineering, physiology, organisational behaviour and very careful writing. Almost nobody enters it young, which is itself the opportunity.
A large telescope's observing schedule is one of the most contested on Earth, and a sky survey is a data pipeline whose quality decides what gets discovered. Scheduling, calibration and control charts, under the sky.
Roughly half the students who take this will never work in space. They will still have spent eight years learning to define done, estimate honestly, see risk early, measure before deciding, and finish. That transfers everywhere, including to being an adult.
09 How it runs
One student, one teacher, on Zoom, at a time that suits a school week and a family. No recordings of somebody else's lesson, no cohort to keep pace with, no teaching assistant.
What was covered, what your child actually understood as opposed to what they said, and what to look for during the week. You are not left guessing whether progress is happening.
Abhishek Dabas, MSc Astronautics and Space Engineering, BEng Aerospace Engineering. Every class personally. A deliberately small practice, which is the reason that is possible.
Because this material cannot be delivered. It has to be argued with. A student who says the build will take three weeks needs to be asked what would make it five, and a student who says the fix worked needs to be asked how they know it was not simply a good week, and each question has to arrive within four seconds of the claim, from somebody who knows which of the four usual optimisms produced it. That does not scale, and I have stopped pretending it does.
10 Questions
Because in real work they are one job. A project is how something gets delivered on a date; Six Sigma is how the thing delivered is right and stays right; every mission has both problems at once and every improvement is itself a project. Teaching them apart is how adults end up with a Gantt chart and no data, or a control chart and no deadline. The nine disciplines on this page are what a programme office and a quality lab actually use, taught together from the first term.
A management course, taught on the space, aviation and astronomy world. Every example, every case and every project comes from there, because that is the industry where both disciplines are used at their most unforgiving: a window set by the planets, and hardware that gets one attempt. A child who loves rockets learns critical paths and control charts without noticing, and the habits transfer to everything else they will ever do.
The vocabulary is corporate. The skills are older than corporations: deciding what finished means, working out the order things must happen in, estimating honestly, imagining what could go wrong, measuring before deciding, and telling people the truth about progress. Children are handed projects constantly and taught none of that. They take to it faster than adults do, because they have not yet learned to hide a late project behind a confident update or defend themselves with opinions.
Grade 5. Stage One runs real projects a ten year old can own, a model rocket launch day, an observing night, a science fair entry, with a plan on one page and a measurement at the end. It then runs to Grade 12 as one continuous plan rather than as three unrelated courses. Students join later than Grade 5 regularly, and the assessment is how I work out where they should start.
All of them, as dialects of one language. The arithmetic of a critical path and the mathematics of a control chart do not change with the manual. Where the frameworks differ, a student learns both versions and, more importantly, learns why they differ, which is a better lesson than any one of them alone.
Yes. PMI's Project Management Ready certification was built for high school students, and the Yellow Belt from the American Society for Quality has no experience or education requirement. At eighteen the CAPM needs only a school diploma and twenty three hours of project management education, which this plan exceeds many times over, and there are Green Belt examinations whose only prerequisite is knowing the material. Stage Three is taught to those bodies of knowledge, so a student who wants a certificate sits it as revision, with the certifying body directly. Certification is a by product here, not the point; the point is the habits.
Exactly as much as the mission needs, at the moment it needs it. Arithmetic, countdowns and averages in Stage One; network arithmetic, probability and the standard deviation in Stage Two; inference, expected value, distributions and simulation in Stage Three. Students here meet probability and statistics a year or two ahead of school without resistance, because the alternative is not knowing whether the window is real or whether the fix worked.
That is the usual case, and no. The material has no artificial ceiling: Stage Three sits at the level of the entry professional certifications and the Green Belt body of knowledge, and a student who wants to go further goes further in the same hour, because there is nobody else in the room to hold to.
I do. Abhishek Dabas, a rocket engineer with an MSc in Astronautics and Space Engineering and a BEng in Aerospace Engineering. Every class, personally, never a group and never a stand in. The practice is deliberately small across all five subjects.
With a seventy five minute assessment. Forty five minutes with your child, working real problems rather than being interviewed, and thirty with you. Within forty eight hours you receive a written assessment of your child's reasoning: what they understand, what they only appear to understand, and what I would do about it. It is yours to keep whatever you decide afterwards.
11 How every family begins
Seventy five minutes: forty five with your child working real problems, and thirty with you. Within forty eight hours you receive a written assessment of their reasoning, yours to keep whatever you decide.
A deliberately small practice · taught personally by a rocket engineer