Read automatically from the scorecard below — every claim can be checked against it.
68 % of the score comes from initial soc error (T10, 20.12 % × weight 0.100) and current sensor offset (T11, 18.43 % × weight 0.100) — improving there moves the leaderboard position most.
Accuracy holds on the never-released cell (0.67 % vs 0.73 % on the open cells) — the model generalises rather than memorising the released data.
Cold is the weak spot: 1.21 % at −20 °C against 0.55 % at 25 °C. Below zero the cell's resistance rises and the voltage curve distorts; models without temperature awareness lose the most here.
Charging is much worse than driving (13.39 % vs 0.71 % overall) — typical of estimators tuned only on discharge behaviour.
A wrong initial SOC is not corrected: 20.12 % on test 10. Pure current integration can never recover from a bad start; voltage feedback is what fixes this.
Scorecard
Average RMSE (% SOC) per test case, weighted into the leaderboard score. Lower is better.
Estimation accuracyT1All cells(reference only — every other test is a subset of it)
0.71 %
0.0000
0.0000
T2Blinded cell (m448)
0.67 %
0.1000
0.0667
T3
Key cases
The runs that separate estimators — cold and hot cycles, the blinded cell, wrong initial SOC, a biased current sensor. Each plot says why it is there.
Selected drive cycles
The thirteen cycles the original MATLAB tool plotted, one per test: blinded vs non-blinded cell (tests 1–3), drive vs charge (test 4), the three payloads on one cycle (tests 5–6), standard vs non-standard highway profiles (tests 7–8) and the temperature extremes (test 9).
m80 UDDS at -20 °C
Whole cycle · RMSE 1.50 %, max 2.7 %
Actual SOC
Dynamic Neural Network.
Error axis
Why this case: Test 9 — coldest condition: high resistance and a strongly non-linear OCV response — the hardest cycle for most estimators.
Wrong initial SOC (test 10)
Three cycles restarted with the estimator told the battery is at 90, 60 or 30 % while it is really full. Coulomb counting can never recover; voltage-informed estimators converge at a rate that is very visible here.
Wrong initial SOC 90 % — m80 LA92 at 25 °C
Whole cycle · RMSE 6.27 %, max 9.0 %
Actual SOC
Dynamic Neural Network.
Error axis
Why this case: The cycle is joined at the point where the true SOC has already fallen to 90 %, so a model that assumes a full battery at start-up begins 10 % off. What matters is how fast the estimate converges onto the truth.
Current-sensor offset (test 11)
Three cycles re-run with ±0.3 A added to the measured current. Integration-based estimators drift steadily; the plot shows whether and how fast the model corrects the bias.
Current-sensor offset -0.3 A — m1000 US06 at -10 °C
Whole cycle · RMSE 11.49 %, max 13.9 %
Actual SOC
Dynamic Neural Network.
Error axis
Why this case: A constant -0.3 A is added to the measured current before it reaches the model. Pure current integration drifts linearly; estimators that also use voltage should correct it.
RMSE vs. sensor offset
Test 11 — the same three cycles re-run with a constant current offset; 0 A is the unmodified run. Only ±0.3 A enters the score.
US06 at −10 °C
HWFET at 10 °C
LA92 at 40 °C
Test-case chartsThe scorecard as bar charts — tests 1–8, and RMSE against temperature (test 9).
Error by test case
Average RMS error (% SOC) for each blinded test case. Lower is better.
Error vs. ambient temperature
Test 9 — m80 cell, all blinded cycles at each chamber temperature.
All 144 cyclesEvery blinded drive cycle: per-cycle errors, and any of the plotted cycles in the time domain.
Cell
Temp.
Cycle
Duration
RMSE
MAE
MAXE
m80
-10 °C
UDDS
6.32 h
0.39
0.27
0.84
m80
-10 °C
HWFET
3.02 h
0.87
0.71
1.44
m80
-10 °C
LA92
4.74 h
0.37
0.30
0.72
m80
-10 °C
US06
2.30 h
0.37
0.30
0.66
m80
-10 °C
HWCUST
1.69 h
0.34
0.27
0.63
m80
-10 °C
HWGRADE
1.24 h
0.36
0.29
0.74
m80
-20 °C
UDDS
4.62 h
1.50
1.17
2.73
m80
-20 °C
HWFET
2.34 h
1.67
1.34
2.70
m80
-20 °C
LA92
3.54 h
0.38
0.29
1.07
m80
-20 °C
US06
1.86 h
2.19
1.61
4.79
m80
-20 °C
HWCUST
1.37 h
1.09
0.89
1.68
m80
-20 °C
HWGRADE
1.25 h
0.42
0.37
0.69
m80
0 °C
UDDS
8.75 h
0.52
0.41
1.04
m80
0 °C
HWFET
3.84 h
0.39
0.34
0.71
m80
0 °C
LA92
6.39 h
0.72
0.53
2.03
m80
0 °C
US06
2.93 h
0.33
0.29
0.57
m80
0 °C
HWCUST
2.08 h
0.26
0.22
0.46
m80
0 °C
HWGRADE
1.64 h
0.65
0.45
1.30
m80
10 °C
UDDS
11.32 h
0.40
0.35
0.77
m80
10 °C
HWFET
4.49 h
0.63
0.58
1.02
m80
10 °C
LA92
7.88 h
0.41
0.35
0.78
m80
10 °C
US06
3.42 h
0.46
0.42
1.11
m80
10 °C
HWCUST
2.42 h
0.21
0.19
0.39
m80
10 °C
HWGRADE
1.80 h
0.71
0.54
1.41
m80
25 °C
UDDS
16.13 h
1.17
0.93
2.57
m80
25 °C
HWFET
5.52 h
0.59
0.54
0.94
m80
25 °C
LA92
10.58 h
0.26
0.21
0.52
m80
25 °C
US06
4.22 h
0.64
0.60
0.91
m80
25 °C
HWCUST
2.86 h
0.45
0.43
0.66
m80
25 °C
HWGRADE
2.33 h
0.20
0.18
0.27
m80
40 °C
UDDS
12.40 h
0.76
0.58
1.81
m80
40 °C
HWFET
5.41 h
0.96
0.87
1.68
m80
40 °C
LA92
9.07 h
0.57
0.52
0.85
m80
40 °C
US06
4.17 h
1.19
1.09
1.96
m80
40 °C
HWCUST
2.95 h
0.71
0.65
1.25
m80
40 °C
HWGRADE
2.73 h
0.58
0.54
0.94
m448
-10 °C
UDDS
6.23 h
0.42
0.28
0.92
m448
-10 °C
HWFET
3.01 h
1.07
0.88
1.75
m448
-10 °C
LA92
4.59 h
0.58
0.50
1.05
m448
-10 °C
US06
2.25 h
0.51
0.39
1.33
m448
-10 °C
HWCUST
1.69 h
0.66
0.57
1.04
m448
-10 °C
HWGRADE
1.22 h
0.29
0.20
0.68
m448
-20 °C
UDDS
4.56 h
1.40
1.10
2.50
m448
-20 °C
HWFET
2.34 h
1.72
1.38
2.77
m448
-20 °C
LA92
3.46 h
0.51
0.43
0.91
m448
-20 °C
US06
1.80 h
1.73
1.00
4.27
m448
-20 °C
HWCUST
1.39 h
1.53
1.30
2.26
m448
-20 °C
HWGRADE
1.23 h
0.74
0.69
1.02
m448
0 °C
UDDS
8.60 h
0.50
0.39
1.03
m448
0 °C
HWFET
3.83 h
0.38
0.33
0.71
m448
0 °C
LA92
6.16 h
0.41
0.33
1.15
m448
0 °C
US06
2.87 h
0.44
0.37
1.10
m448
0 °C
HWCUST
2.09 h
0.34
0.29
0.62
m448
0 °C
HWGRADE
1.51 h
0.60
0.42
1.30
m448
10 °C
UDDS
11.08 h
0.42
0.38
0.70
m448
10 °C
HWFET
4.47 h
0.69
0.63
1.09
m448
10 °C
LA92
7.60 h
0.56
0.48
1.04
m448
10 °C
US06
3.35 h
0.69
0.65
1.07
m448
10 °C
HWCUST
2.42 h
0.23
0.15
0.62
m448
10 °C
HWGRADE
1.68 h
0.88
0.67
1.72
m448
25 °C
UDDS
15.59 h
0.32
0.27
0.68
m448
25 °C
HWFET
5.46 h
0.66
0.59
1.14
m448
25 °C
LA92
10.06 h
0.26
0.23
0.49
m448
25 °C
US06
4.07 h
0.49
0.31
1.78
m448
25 °C
HWCUST
2.86 h
0.36
0.34
0.56
m448
25 °C
HWGRADE
2.19 h
0.17
0.16
0.24
m448
40 °C
UDDS
12.09 h
0.76
0.69
1.36
m448
40 °C
HWFET
5.21 h
1.18
1.06
2.13
m448
40 °C
LA92
8.79 h
0.73
0.65
1.05
m448
40 °C
US06
4.06 h
0.72
0.68
1.29
m448
40 °C
HWCUST
2.92 h
0.58
0.53
1.01
m448
40 °C
HWGRADE
2.32 h
0.47
0.44
0.70
m448-N
-10 °C
UDDS
10.80 h
1.42
1.18
2.38
m448-N
-10 °C
HWFET
3.85 h
0.35
0.27
0.72
m448-N
-10 °C
LA92
6.66 h
0.50
0.39
0.92
m448-N
-10 °C
US06
2.70 h
0.69
0.36
2.01
m448-N
-10 °C
HWCUST
1.89 h
0.32
0.25
0.63
m448-N
-10 °C
HWGRADE
1.36 h
0.60
0.47
1.17
m448-N
-20 °C
UDDS
8.67 h
1.98
1.83
2.71
m448-N
-20 °C
HWFET
3.09 h
1.76
1.39
2.97
m448-N
-20 °C
LA92
5.21 h
0.43
0.40
0.95
m448-N
-20 °C
US06
2.23 h
4.14
3.10
8.82
m448-N
-20 °C
HWCUST
1.65 h
1.47
1.20
2.28
m448-N
-20 °C
HWGRADE
1.50 h
0.46
0.41
0.74
m448-N
0 °C
UDDS
12.79 h
1.73
1.48
2.71
m448-N
0 °C
HWFET
4.47 h
0.16
0.13
0.34
m448-N
0 °C
LA92
7.99 h
1.04
0.84
2.08
m448-N
0 °C
US06
3.22 h
1.00
0.53
3.07
m448-N
0 °C
HWCUST
2.31 h
0.22
0.16
0.58
m448-N
0 °C
HWGRADE
1.65 h
0.74
0.57
1.40
m448-N
10 °C
UDDS
13.92 h
0.81
0.64
1.61
m448-N
10 °C
HWFET
4.84 h
0.49
0.45
0.84
m448-N
10 °C
LA92
8.80 h
0.36
0.30
0.74
m448-N
10 °C
US06
3.56 h
0.33
0.30
0.56
m448-N
10 °C
HWCUST
2.53 h
0.39
0.27
0.94
m448-N
10 °C
HWGRADE
1.79 h
1.15
0.91
2.12
m448-N
25 °C
UDDS
15.61 h
0.47
0.39
0.92
m448-N
25 °C
HWFET
5.47 h
0.28
0.25
0.50
m448-N
25 °C
LA92
10.09 h
0.33
0.26
0.76
m448-N
25 °C
US06
4.08 h
0.71
0.41
2.56
m448-N
25 °C
HWCUST
2.86 h
0.16
0.14
0.30
m448-N
25 °C
HWGRADE
2.30 h
0.17
0.14
0.33
m448-N
40 °C
UDDS
16.75 h
1.55
0.90
3.86
m448-N
40 °C
HWFET
5.89 h
0.93
0.84
1.71
m448-N
40 °C
LA92
11.03 h
0.59
0.53
1.22
m448-N
40 °C
US06
4.44 h
0.61
0.58
0.83
m448-N
40 °C
HWCUST
3.13 h
0.50
0.46
0.91
m448-N
40 °C
HWGRADE
2.48 h
0.21
0.19
0.36
m1000
-10 °C
UDDS
5.91 h
0.27
0.20
0.60
m1000
-10 °C
HWFET
2.91 h
1.00
0.82
1.62
m1000
-10 °C
LA92
4.23 h
0.42
0.36
0.76
m1000
-10 °C
US06
2.09 h
1.21
0.70
3.27
m1000
-10 °C
HWCUST
1.67 h
0.72
0.62
1.13
m1000
-10 °C
HWGRADE
1.20 h
0.38
0.29
0.74
m1000
-20 °C
UDDS
4.31 h
0.73
0.61
1.31
m1000
-20 °C
HWFET
2.27 h
1.75
1.41
2.83
m1000
-20 °C
LA92
3.20 h
1.15
0.71
2.95
m1000
-20 °C
US06
1.74 h
1.81
1.68
2.31
m1000
-20 °C
HWCUST
1.37 h
1.43
1.20
2.13
m1000
-20 °C
HWGRADE
1.22 h
0.38
0.34
0.59
m1000
0 °C
UDDS
8.09 h
0.30
0.24
0.64
m1000
0 °C
HWFET
3.70 h
0.48
0.43
0.82
m1000
0 °C
LA92
5.60 h
0.74
0.52
2.29
m1000
0 °C
US06
2.66 h
0.40
0.37
0.79
m1000
0 °C
HWCUST
2.05 h
0.44
0.38
0.77
m1000
0 °C
HWGRADE
1.36 h
0.68
0.49
1.41
m1000
10 °C
UDDS
10.51 h
0.45
0.34
0.96
m1000
10 °C
HWFET
4.37 h
0.74
0.68
1.14
m1000
10 °C
LA92
7.06 h
0.58
0.54
0.93
m1000
10 °C
US06
3.11 h
0.66
0.62
0.96
m1000
10 °C
HWCUST
2.41 h
0.28
0.16
0.74
m1000
10 °C
HWGRADE
1.52 h
1.14
0.85
2.16
m1000
25 °C
UDDS
14.69 h
0.32
0.26
0.70
m1000
25 °C
HWFET
5.37 h
0.68
0.61
1.11
m1000
25 °C
LA92
9.30 h
0.50
0.33
1.93
m1000
25 °C
US06
3.79 h
1.12
0.66
3.97
m1000
25 °C
HWCUST
2.83 h
0.50
0.48
0.72
m1000
25 °C
HWGRADE
1.90 h
0.20
0.18
0.29
m1000
40 °C
UDDS
11.69 h
0.71
0.64
1.04
m1000
40 °C
HWFET
5.15 h
1.11
1.01
1.95
m1000
40 °C
LA92
8.33 h
0.60
0.56
0.85
m1000
40 °C
US06
3.89 h
0.78
0.61
2.27
m1000
40 °C
HWCUST
2.91 h
0.68
0.62
1.16
m1000
40 °C
HWGRADE
2.17 h
0.62
0.57
1.05
m80 UDDS at -20 °C
Whole cycle · RMSE 1.50 %, max 2.7 %
Actual SOC
Dynamic Neural Network.
Error axis
One hour of padded data precedes every cycle in the evaluator and is excluded from the error metrics. Charts are down-sampled for display (peaks preserved); the full 1 Hz data is in the traces download below.
Score historyEvery evaluation attempt and every change to how this submission is scored; the current score is the last row.
Aug 31, 2026, 03:27 a.m. EDTEvaluation failed · realattempt 1: Evaluation exceeded the 180 minute limit.failed
How a other worksThe canonical structure for this model family; the author's description below gives the specific architecture.
Other — standardized view
Unspecified
Governing equation
[Y,z]=Model(X,z)
Downloads
PDF report (summary, every test case with the arithmetic, key plots, per-cycle table) · results as JSON · full 1 Hz traces of all 195 runs as a MATLAB v7 file (0.01 % SOC steps; readable with scipy.io.loadmat, see the readme variable inside).
Torino_Federico_Updated_DYNN_Fast.zip (15.5 KB) — deleted after evaluation
Visibility
Public
Evaluator
socbench-eval-0.1.0/matlab
Completed
Aug 31, 2026, 10:52 a.m. EDT
Submission ID
cmtgqm0j80005jl041levjtd9
Non-blinded cells
0.73 %
0.1000
0.0729
Operating conditionsT4Charging
13.39 %
0.1000
1.3395
T580 kg payload
0.66 %
0.0333
0.0219
T5448 kg payload, HVAC on
0.67 %
0.0667
0.0445
T6448 kg payload, HVAC off
0.81 %
0.0667
0.0538
T51000 kg payload
0.72 %
0.0333
0.0240
T7Standard drive cycles
0.79 %
0.1000
0.0792
T8Non-standard drive cycles
0.56 %
0.1000
0.0556
T9−20 °C
1.21 %
0.0167
0.0202
T9−10 °C
0.45 %
0.0167
0.0075
T90 °C
0.48 %
0.0167
0.0079
T910 °C
0.47 %
0.0167
0.0078
T925 °C
0.55 %
0.0167
0.0092
T940 °C
0.80 %
0.0167
0.0133
RobustnessT10Initial SOC error
20.12 %
0.1000
2.0121
T11Current sensor offset
18.43 %
0.1000
1.8427
Weighted error = Σ (weight × RMSE)
5.68 %
Max error 8.8 % — largest instantaneous |estimate − truth| over every blinded cycle (on m448-N US06 at -20 °C — plotted under Key cases for new evaluations); not scoredComplexity 10 ±1 · Extreme+ (informational, not scored)
Ahnaf Rahman · McMaster University