Most FTC teams do not fail because of a bad robot. They fail because of lost knowledge. When a strong senior cohort graduates, it leaves with the CAD shortcuts, the wiring habits, the autonomous code and the competition instincts that took two or three seasons to build. The team that returns in the new year carries last year's name and a rookie's memory. The way out is to treat succession as an engineering problem you design for from the first meeting, not an emergency you notice in the off-season. In practice that means five habits: documenting relentlessly, pairing every senior with an understudy, spreading leadership across sub-teams, recruiting a year early, and protecting coach continuity.

Why teams stall the year after a big cohort leaves

The pattern is familiar. One or two capable students hold most of the working knowledge in their heads. They write the autonomous routine, they know which bolt strips if you overtighten it, they remember why the intake was rebuilt in week six. Nobody wrote any of it down, because the people who knew did not need notes. Then they matric, and the knowledge walks out the door with them. The rebuild in the new year is not really a rebuild. It is a reinvention, and the team spends its first months relearning lessons it already paid for.

Sustainability is the antidote, and it is a design choice. A team that expects to lose a quarter of its brains every year builds so that losing them hurts less.

Make documentation a team habit, not a chore

The engineering notebook is not just a judging requirement. It is the team's memory. Treat it as a living record that anyone could use to rebuild a subsystem from scratch. Write the "why" behind decisions, not only the "what": why you chose the drivetrain, why you abandoned a mechanism, what failed at which event. Keep a shared drive with CAD files, wiring photographs and a one-page build sheet for every subsystem. Comment the code so a student who has never seen it can follow the autonomous logic.

Give rookies a low-stakes place to practise so the notes have somewhere to land. A simple microcontroller board like the sheenbot∞ lets newer members learn sensors, loops and motor control without ever touching the competition robot, so by the time they read the team's code it already makes sense.

Give every senior an understudy

Shadow roles are the single most effective succession tool a small team has. Pair each senior specialist with a younger student for a full season. Early on, the senior leads and the understudy watches. By mid-season the understudy does the work while the senior watches and corrects. By competition time the younger student can run the task alone if they have to. The goal is plain: no skill on the team should live in exactly one person's hands. If a student ever says "only I know how to do this", that is not a badge of honour, it is a risk to fix before they leave.

Spread leadership across sub-teams

A single all-knowing captain is a single point of failure. Break the team into sub-teams with named leads: mechanical, programming, electrical, outreach and documentation. Each lead owns their area and trains their own successor. When the seniors go, graduation removes a slice of the team rather than its whole brain. Distributed leadership also grows more leaders, because five students get to run something instead of one, and the students left standing next season have already led before.

Recruit a year early and build a feeder

The teams that survive graduation are the ones that recruited the replacements before they were needed. Do not wait for a gap to appear. Bring grade 8 and 9 students in as apprentices while the seniors are still around to train them. Run an open build night, visit younger classes, and use the school holidays as an on-ramp. Holiday robotics workshops are a natural feeder: students who spend a week building and coding arrive already curious and half-trained. A younger cohort that shows up with some grounding in coding means less time spent on the basics and more spent on the robot.

Protect coach and mentor continuity

Succession is not only about students. Coaches move schools, burn out, or hand over. Document the team's operating knowledge too: sponsors, registration deadlines, travel logistics, budget and the season calendar. A co-coach model, where two adults share the load, means the team survives one of them stepping back. Parent mentors and returning alumni matter here. An alum who comes back to mentor carries forward exactly the institutional memory you are trying to keep.

An off-season succession checklist

• Every subsystem has a one-page build sheet and photographs on the shared drive.
• The autonomous and tele-op code is commented well enough for a newcomer to follow.
• Each senior has a named understudy who has done the job unsupervised at least once.
• Sub-team leads for next season are assigned before this one ends.
• At least two younger students are recruited and partly trained per graduating senior.
• Coach knowledge (sponsors, deadlines, logistics) is written down, not just remembered.
• A spare practice base or training kit exists so rookies learn without risking the competition robot.

The takeaway

A team that only wins when its best students are present is fragile. A team that has documented its knowledge, trained understudies, shared its leadership and recruited early can lose a cohort and come back steady. Build those habits into the ordinary rhythm of the season, not into a panicked scramble the term before matric exams, and graduation becomes a handover instead of a collapse. For more coaching and build notes, keep an eye on our newsroom.

Frequently asked questions

How early should we start planning for succession?

From a student's first season, in small ways. The most important window is the season before a big cohort graduates. If your seniors leave at the end of this year, their understudies should already be doing real work now, while there is still someone to correct them.

What if only one student understands our code?

Treat that as an urgent risk, not a convenience. Have that student write a short walkthrough of the autonomous logic and pair-programme with at least one younger member every session. Move the code to a shared repository with comments, so it is not trapped on one laptop or in one head.

How do we keep a coach from term to term?

Share the load and lower the barrier. A co-coach or a small rotation of parent mentors is far more durable than one heroic volunteer. Write down the logistics so a new coach can step in without reverse-engineering a year of decisions, and invite alumni back as mentors once they leave.