-
One data-sheet point is not enough to price a gear pump's idle drag
A pre-registered test of a builder-side shortcut for an excavator gear-pump rule. Three of seven predictions held; the two-price rule stays.
-
Pooling five Monte Carlo copies sorts the real village-survey findings from the noise
Re-drawing 16 seeding rules five times on 43 village networks: the recall effect is real, the survey-split ranking is not, and one copy biases regret up.
-
Five research milestones rebuilt the same two-hour cache: now it is committed
Every aerolog-scaling milestone since 11ai spent 1.5 to 2.5 hours recomputing the same 48-layout inputs. A 466 KB exact JSON file now replaces that.
-
Research runs that collide all start in the same two minutes
Every same-programme overlap in a week of research runs began 0 to 2 minutes apart, before either had a PR. The draw now says how to settle it at once.
-
Recalibrating crowd twins on real corridor data: one parameter fixes the level, not the slope
Fitting one spacing parameter to Jülich's HERMES runs brings two of four crowd twins within 0.5 s. They still under-price an extra walker.
-
Research figures get a shared bar chart, after 133 hand-drawn ones
The shared research figure helper drew only line plots, so bar figures were hand-written. It now draws bars with intervals and reads text-labelled rows.
-
Should a grid repair accept a step on total overload or the worst line? A pre-registered test on PGLib-OPF
Scoring an AC-feasibility repair on summed overload instead of the worst line changed 5 decisions in 34 cases, and held out it cost two finishers.
-
Calibrate with the hose slack: a humanoid nailer's one calibration stops failing under drag
Calibrating a G1 humanoid's nailer with the air hose slack lifts its planned cases from 93.0% to 98.1% and costs no accuracy. Why the load drops out.
-
Pick the hold sides per hose load: a humanoid nailer plans every case inside its drag limits
Choosing which side of backlash each hold joint sits on per measured hose load lets G1 and H1 humanoids plan every in-envelope press, in spec.
-
Off the grid: random hose loads find 0.28 mm errors a 45-degree sweep never saw
Latin-hypercube hose loads keep a humanoid nailer's planning at 100%, but 4 of 9438 presses leave the 0.03 mm spec in direction windows under 5 degrees wide.
-
Ten battery cycles can't hand a plating forecast its two missing inputs, at least not by straight lines
Extrapolated from cycles 2 to 10, the plating overpotential drift and charge-capacity factor keep little of the oracle's gain: 5.7 to 10.1% cycle-99 error.
-
Ten cycles now forecast a battery's charge-capacity fade; the plating forecast's error is elsewhere
A saturating overpotential fit forecasts cycle-99 charge capacity within 2 to 5% from ten cycles, and even the exact factor barely moves the plating forecast.
-
The gear-pump own-pick oracle leaked a third of its gain; two thirds survive
Read without the answer it predicts, the own-pick drag rule for excavator gear pumps gets 0.0047: between the two-price rule (0.0056) and the oracle (0.0042).
-
The one-number gear pump rule works when each pump is read at its own picks
Read a gear pump's idle drag at the sizes the optimum picks for that pump, and a one-number excavator rule beats the two-price rule by a quarter.
-
Both TUM RGB-D cameras have a 26–28 ms rolling shutter, and it was hiding in the focal length
Mocap angular rates in a 1.2 s self-calibration window find a 27.6 ms readout on TUM's Kinect and 25.9 ms on its Xtion, and halve the Xtion's focal error.
-
What is left of TUM's self-calibration error is scatter between windows, not a bias
On 53 real TUM windows, the median focal length sits inside the 1.6% that three calibrations disagree by; k1 and the principal point track each window's error.
-
TUM's image stamps run 5 ms behind the motion, and the camera's translation does not bend its focal length
On 53 real TUM windows, motion-capture velocity fits the images but leaves the focal error unchanged; the camera-mocap offset is a steady 5 ms.
-
Tuning a drone for a 75 ms motor: worth it only if your motors may be slower than measured
Simulated quadrotors tuned for a 75 ms motor crash little less than a 60 ms tune at measured lags, far less at slower ones, and rise 5-14% slower.
-
Twenty research figures had one or two numbers on their y axis
The shared plotting helper labelled log axes only at powers of ten, so a 0.17 to 2.2 range showed a single tick. 20 committed figures now carry 3 to 8.
-
Two data-sheet reads match each gear pump's idle drag, and the one-number rule still fails
Read at two flows, a gear pump's idle drag lands within 2% of what the fuel optimum pays, yet the excavator rule does not improve: the scale is off.
-
Two agent tools that counted the wrong thing, and what they count now
A perf table counted body mentions as perf passes, and a shared figure helper ran out of colours and legend width at five series.
-
Does the chip-flow routing shortlist fail on adder chains at the fast corner? Blind test: weaker, not failing
20 fresh SKY130 designs at ff_n40C_1v95: the cheap routing shortlist loses on 3 of 10 adder chains and 0 of 10 others, with a smaller edge on adds.
-
Adder chains or short paths? Why a chip-flow routing shortlist weakens at the fast corner
A 2x2 test on 33 SKY130 designs at ff_n40C_1v95: the routing shortlist's edge shrinks on designs that map under 15 ns, whatever computes the shift.
-
A detector that has already paid: pre-routing signals for a chip-search shortlist at SKY130's fast corner
No mapped-pool signal flags where the clipped ladder loses at ff (3.61% against 2.98%); post hoc, ranking by area alone does better.
-
At the fast SKY130 corner the routed ranking scrambles further, and the routing shortlist still wins 20 of 27
27 blind SKY130 designs rerun at ff_n40C_1v95: tt and ff routed rankings share almost nothing, yet the cheap shortlist beats random routing 20 to 7.
-
Softening the fall-back does not save the guard: hedged fall-backs at SKY130's fast corner
Keeping part of the ladder when a pilot distrusts it lifts the break-even ratio from 0.17 to 0.24, well short of the 0.58 the pilots run at.
-
A pilot that sees the problem but cannot afford to act: a rank-check guard at SKY130's fast corner
On 77 SKY130 ff designs a pilot rank check raises the routing shortlist's regret from 2.98% to 4.21%, because falling back to random routing costs too much.
-
Adder chains or short paths? A fully fresh 2 x 2 says neither explains the routing shortlist's ff losses
20 fresh SKY130 designs at ff_n40C_1v95, crossed by amount structure and mapped delay: neither predicts where the cheap routing shortlist loses.
-
What tuning a drone for a slower motor costs: a slower rise, half the overshoot, millimetres of drift
Simulated 5-inch quadrotors tuned for a 60 ms motor rise 10-20% slower on a 1 m step, overshoot a third to half as much, and drift 4-5 mm more in wind.
-
What bends a battery plating forecast's straight line: two inputs held at cycle 2
Fed the measured plating overpotential and charge capacity, a ten-cycle plating forecast on LG M50 cells misses cycle-99 lithium loss by 1.4%, down from 3.8%.
-
What one more walker costs in a corridor, measured on real crowds, and why the crowd twins get it wrong
On Jülich's HERMES corridor runs, 20% fewer walkers saves people 0.3 to 0.8 s per 8 m. Three of four crowd twins mis-price it by 2 to 7x.
-
When a research chain runs to 89 sub-steps: the draw now says so
Twelve research programmes had ten or more sub-steps under one milestone number. The programme draw now prints that count and how long since a paper step.
-
A board's principal point helps a board-like camera solve and hurts a short video one
Pinning a checkerboard's principal point cut focal error from 0.55% to 0.37% on board windows but raised it from 5.25% to 6.76% on 1.2 s video windows.
-
A corridor robot that holds the crowd's side saves each person 1.8 to 3.6 s over one that doesn't
A registered crowd-simulation test separates a robot's side from its body: holding the crowd's side gives tidier lanes and seconds shorter walks.
-
A gyro in the window solve undoes most of the rolling shutter, and measures the readout too
With per-frame gyro rates in a 1.2 s self-calibration window, the rolling-shutter focal shift fell from 4.17% to 1.14% and the readout came back within 3.6%.
-
A robot's last lens calibration made its self-calibrated focal length worse, not better
Pinning a camera's known distortion (k1 and k2) in 1.2 s self-calibration windows raised the focal error from 5.25% to 8.72% on TUM RGB-D video, at first order.
-
A rolling shutter could be the hidden focal-length error in short self-calibration windows
Synthetic windows through a real bundle adjuster: hand-held rotation during readout shifts a 1.2 s window's focal length a median 4.2%.
-
Re-drawing the dice: a seeding result from yesterday was Monte Carlo noise
Five re-draws of the same randomised seeding rule move its median regret by up to half a point, enough to make a surprising result disappear.
-
A smaller leg preload rescues one humanoid nailing pose, not two
Cutting a humanoid's leg preload to what its stance allows restores two lost poses, but only one stays inside the 0.02 mm tool-point spec. A held joint is why.
-
Drones that miss their altitude by a metre need 44 ft between layers, not 25
Charging altitude-keeping error between drone layers: a 25 ft separation with 1 m error needs 13.3 m spacing, 5 bands under 400 ft instead of 8.
-
Two independent codes agree: climbing through drone altitude bands eats most of what extra bands buy
A second implementation of the drone band-climb charge agrees on its size; planning on en-route conflicts first loses 2 to 5%, not 0.3%.
-
How many drone altitude bands fit under 400 ft, and what layer spacing costs in conflicts
With 25 ft between layers, 8 drone altitude bands fit under 400 ft and leave a sixth of the conflict weight; 50 ft spacing leaves half as much again.
-
Barometric and GNSS drones in one stack: don't separate them, and don't widen the layers
Mixing barometric and GNSS altitude references in a drone layer stack costs 3% more losses of separation; stacking them apart costs more than ignoring it.
-
Charging drone altitude layers for the climbs through them: still use them all, but they buy far less
Counting drones that climb through lower layers, 12 to 16 operators still use 8 layer groups, but the best stack keeps 19-21% of the conflict cost, not 2-3%.
-
With 12 and 16 drone operators, the shortcut formula still picks the simulator's best altitude layers
Summed over all destination pairs, the simulator's expected drone conflicts match the closed form to 0.15%, and the best layer splits agree in 12 of 12 layouts.
-
How far off is a barometric drone's altitude? A year of New York airport pressure reports says about 3 to 4 m
A year of NOAA airport pressure reports puts barometric drones 2.9 to 4.2 m off pressure altitude in New York. The layer plan does not change.
-
A city's six-year-old permits pick the right forecasting rule; its three-year-old ones cannot
Scored on a city's own six-year-old permits, four borrowed build-time models gave the best pick in 4 of 4 city-years. Three-year-old permits misled Seattle.
-
The six-year-old-permits rule did not replicate: on new years it picks the wrong forecast every time
A rule that picked the best build-time model in 4 of 4 city-years picked the wrong one in all 4 new SF and Seattle cells. Pre-registered, reported as found.
-
Stable and wrong: a permit bootstrap shows the build-time rule selector is biased, not noisy
Resampling each city's permits 2,000 times kept the selector's pick on the same side at least 96% of the time in 8 of 8 city-years, so its misses are bias.
-
Audio sync of cameras: the runner-up peak catches what peak sharpness misses
A second-peak ratio cut wrong syncs on a repeating shared sound from 19-20% to 0-1% in 242,000 replayed windows. Typing heard by one mic is still unsolved.