Feature
Progressive Overload Tracking
Progressive overload is the foundation of strength training: gradually increasing the demand on your muscles over time. Plates tracks this automatically and tells you exactly when to add weight, how much, and why.
How it decides you're ready
The suggestion engine analyzes your last 20 sessions per exercise. It's not a simple "hit 3x8, add weight" rule. It looks at the full picture:
Track reps + RPE
Log your working sets normally
Learn your profile
ML builds your grind threshold
Detect readiness
Stable effort + consistent reps
Suggest increase
Equipment-aware weight jump
Nine suggestion types
Progressive overload is just one signal. The engine generates nine distinct recommendation types based on your data:
| Suggestion | What it means | When it triggers |
|---|---|---|
| Weight increase | Ready to add load | Hit rep target at stable RPE for 2+ sessions |
| Rep increase | Add reps before weight | Below rep target but effort is low |
| Deload | Time to recover | e1RM dropped 2%+ (personalized threshold) |
| Fatigue warning | Back off slightly | RPE rising without load change |
| Plateau detection | Same weight/reps 4+ sessions | No progress at stable effort |
| Hidden progress | e1RM climbing quietly | Gaining strength you can't see in raw numbers |
| Too light | RPE consistently low | Effort below target zone for 3+ sessions |
| Volume hold | Stay steady | Recent increase, need to consolidate |
| E1RM gain | Estimated max improved | Significant jump in calculated strength |
Real scenarios
Plateau detected
Bench Press: 135 lbs x 8 for 20 sessions with stable effort
You've been hitting 3 sets of 8 at RPE 7.5 for weeks. The engine flags this as a plateau and suggests adding a heavier top set (150 lbs x 4) before your working sets to break through.
Hidden progress
Squat: same weight, but 1RM estimate up 8% over 6 weeks
Your working weight hasn't changed, but your rep quality improved: more reps at the same RPE. Your estimated 1RM is climbing even though you can't see it in the raw log. Plates catches this.
Context-aware hold
Mid-cut: suppressing progression to protect muscle
You're in a calorie deficit. The engine knows your diet phase and holds back weight-increase suggestions. Maintaining load during a cut is the real win, and Plates tells you that.
Equipment-aware weight jumps
Plates rounds suggestions to what you can actually load:
- Barbell: 5 lb / 2.5 kg increments (standard plate pairs)
- Dumbbell: 2.5 lb / 1.25 kg increments
- Cable/machine: 5 lb / 2.5 kg increments (pin-loaded stacks)
It also respects your gym's configured plate inventory, so it never suggests a weight you can't actually build.
Context-aware intelligence
Every suggestion adjusts based on your current state:
- Diet phase: more conservative during a cut, more aggressive during a bulk
- Program phase: suppresses progression during deload/test weeks
- Session readiness: suppresses all progression on a rough day
- Training history depth: needs 3+ sessions before generating any suggestion
On-device, always private
All analysis runs locally on your phone. Your training data never leaves your device for suggestion computation. No cloud ML, no data harvesting, no internet connection required for suggestions to work.