Building, Assessment, and Simulation Explorer for Bike Features
A physics simulation engine for designing and assessing bike jumps, drops, and rollers.
BASE-B generates geometry grounded in real G-force limits, computes risk metrics
like Equivalent Fall Height, and exports blueprint-scale DXF and PDF files —
so builders spend less time moving dirt…Base-Bf is a baseline for the art.
RELEASE SET FOR SPRING OF 2026.
3 feature types.
1 physics engine.
BASE-B handles jumps, drops, and rollers — each with distinct geometry, distinct G-force policies, and distinct construction deliverables. The underlying physics pipeline is nearly identical for all three. Expect more features in v2.
Clothoid ramp transition options, optional gap jumps and step-up and step-down landings, as well a physics-generated constant-EFH landing for saving dirt. A variable EFH landing for a more forgiving profile is also an option, and a surface first developed by BASE-B’s creator as far as we know. Ramp geometry scales automatically to G-force limits by bike mode. Lip height minimum is enforced — the UI won’t let you build a ramp that pinches riders at the angle and speed you specified. There is also a non-EFH tangent landing option for step-ups.
Flat or descending lip, no clothoid ramp, standard ballistic flight and EFH landings. Departure angle from 0° to −20°, optional lip radius for smooth rollover, same three-speed trajectory overlay as jumps. The landing is still physics-generated, not a guess.
Continuous crest-trough profile with G targets at both crest and trough. Liftoff speed computed and annotated. Three speed overlays show whether your design speed is in the desired window, below it, or sending riders airborne for doubles or triples…5-roller chains, up to 25 in a row.
The bike/equipment mode you select sets the G-force caps, takeoff geometry type, wheelbase default, and EFH target threshold. Designing for the largest likely wheelbase in each category is conservative — a feature designed for a 52-inch wb bike works for everything smaller. Users can set a custom wheelbase…up to 8 ft for aMTB.
Continuous clothoid lip — curvature all the way to departure. The curved lip is what gives DJ and BMX riders their pop. Tight G caps reflect compact geometry and high-G tolerance.
Ramp G cap: 2.5 g
EFH PASS threshold: ≤ 3.0 ft
Geometry type: Continuous clothoid, default γ = 0.33 (no flat stab. — adjustable via gamma slider)
Clothoid transition into a flat stabilization section — ~0.25 seconds of uncurved lip before launch. Predictable, calm departure. The flat section is where trail jumps feel “floaty” rather than snappy. But a CCC-no flat can be used if desired.
Ramp G cap: 1.5 g
EFH PASS threshold: ≤ 3.0 ft
Geometry type: Clothoid + flat stab.
Same clothoid-flat geometry as trail/enduro, longer wheelbase floor, slightly higher G cap. DH riders are moving faster and landing harder by design — the constraints reflect that.
Ramp G cap: 2.0 g
EFH PASS threshold: ≤ 3.0 ft
Geometry type: Clothoid + flat stab.
Same clothoid-flat geometry as trail/enduro, with a much longer wheelbase floor reflecting adaptive bike dimensions(user sets the wheelbase). No pump/pop assumed by default — all energy comes from gravity and entry speed. Higher rolling resistance (2× standard) accounts for additional wheels and extra contact patches.
Ramp G cap: 1.5 g
EFH PASS threshold: ≤3.0 ft
Geometry type: Clothoid + flat stab.
Rolling resistance (Crr): 0.04 (2× standard — extra wheels, wider contact)
Pump efficiency: 0% default (no pump; user-adjustable if upper-body pump viable)
One spine.
No alternate paths.
Every run follows a deterministic pipeline — inputs validated, QC checked, surfaces built, flight solved, landing analyzed, outputs assembled. Identical inputs always produce identical outputs. No randomness, no hidden state.
Speed, angle, wheelbase, gap, site grades, and G targets. Imperial or metric — your choice at the top of the UI. All internal physics runs in SI, then convert; the unit toggle is display-only and changes nothing about the calculation.
Profile validation before any physics runs. NaN, non-monotonic coordinates, sparse sampling, and anchor coverage checks. FAIL blocks the pipeline with a clear error. WARN lets it proceed with a visible flag. Nothing slips through silently.
Ramp profile generated from clothoid geometry to your G cap and lip height. Landing built by solving the constant or variable EFH ODE for your design speed plus buffer. One authoritative geometry per run — no divergence between what plots show and what the physics uses.
Full trajectory solved at design speed, design − buffer, and design + buffer. All three trajectories stored and plotted. The slow rider shows where your landing starts; the fast rider shows where it needs to end. Gap clearance for speed setting. A “pop” factor is available for a fourth speed option. Wind envelope overlays headwind and tailwind trajectories as a shaded band — diagnostic visualization showing where wind-shifted landings fall on your surface.
Touchdown location, impact angle, and Equivalent Fall Height computed along the full landing domain. EFH grade (PASS / WARN / FAIL) set by bike mode thresholds. Wheelbase circle drawn at touchdown — showing the physical bike footprint at the moment of landing.
All results assembled into a single auditable output object. Blueprint-scale plots, DXF with labeled layers, and single-page PDF export. Every run is fully reproducible from its RunPack alone — no re-running physics to reconstruct a result.
Measurable. Defined.
Not guessed.
Every output BASE-B produces is grounded in a definition that can be stated, traced, and reproduced. Here are the two that drive the most consequential decisions.
EFH translates landing impact into a number builders and riders can reason about: the height of a flat-ground drop that would produce the same normal impact velocity. A landing with EFH of 2 ft feels like dropping 2 ft straight down — no matter where on the landing you touch for a Constant-EFH surface. BASE-B designs landings to hold EFH constant or variable across the full speed window. The ski industry targets under 4.9 ft (1.5 m); BASE-B’s PASS thresholds are more conservative, set per bike mode. The Variable-EFH surface design is a TrailACT exclusive, born to diversify shapes past Constant-EFH.
G-force in the ramp transition is the difference between a floaty, pump-able jump and one that bucks riders, possibly off the bike. BASE-B’s clothoid geometry keeps centripetal acceleration within the G cap you set by bike mode — 1.5 g for trail/enduro, 2.5 g for DJ/BMX — so the ramp feels smooth at design speed rather than jarring. Landing G thresholds are separate and reflect what riders can typically absorb at touchdown.
A single EFH surface cannot satisfy both headwind and tailwind simultaneously. They’re geometrically incompatible — headwind wants a steep short surface, tailwind wants a shallow long one. Wind analysis is more useful as a diagnostic than a direct design driver because the two constraints are fighting each other geometrically. A surface that’s steep enough to keep EFH low at the far end wants to be shallow near the lip to catch the short case softly, but “shallow near lip” means a flatter initial slope which tends to push the near landing point further out, not closer in.
BASE-Bf’s Variable and Constant EFH surfaces are NOT useful for BOTH tail and headwind at once, but maybe one direction or the other. BASE-B has a headwind and tailwind trajectory band that shows you the problem, you make the judgment call.
The wind envelope is a “how bad is wind at this site” visualizer, not a “fix it” button. If you have a strong prevailing wind you can lean into one of them and decide what is best yourself.
From inputs to
construction deliverables.
BASE-B is built for builders, not just analysts. The outputs are designed to move from screen to ground with minimal interpretation, and freedom to roam — whether you’re working with a tape and a stringline, and a line level, or laser, or a GNSS or RTK rover. The difference: GNSS vs. RTK. Also see: Network RTK vs Traditional Base Stations. Emlid’s RTK: Reach RX2
- Six segmented plot views: run-in, takeoff, gap, landing, takeoff to landing, and full hero layout
- Three trajectory overlays on every jump and drop plot — slow, design/speed setting, and fast rider
- Touchdown wheelbase circle at the design trajectory landing point
- EFH grade callout: PASS (green) / WARN (amber) / FAIL (red)
- Major and minor grid with equal aspect ratio — 1 ft in x equals 1 ft in y, no exceptions
- Gap clearance speed for 1 wheelbase behind speed setting
- Imperial or metric display
PDF build sheet
- Landscape, two pages per feature
- Page 1: full-width side profile with construction-unit grid (1 ft / 3 in imperial, 1 m / 0.2 m metric), dimensioned callouts to lip / knuckle / touchdown / land-end / gap edges, plus a metrics block (geometry, widths, site grade, cut/fill volumes)
- Page 2: stakeout table — Station, x from lip, Elevation, Grade % at user-selectable spacing (auto 1 ft for ≤50 ft features, 2 ft above)
- Walk the feature with a tape: at every station, the table gives the stake’s offset from the lip — horizontal distance and elevation above or below. On hillsides, two extra columns let you lay it out off a stringline run down the natural grade — no transit or laser needed.
DXF
- Named layers: PROFILE, TRAJECTORY, TOUCH_SURFACE, MARKERS, DIMENSIONS, GRADE
- JSON sidecar with every key dimension and design parameter for downstream automation
- For surveyors and CAD operators who’ll dimension in their own tools
CSV
- Raw geometry — every solver point in SI meters, organized by curve (profile, trajectories, touch surface)
- Drop into a spreadsheet for custom analysis, or feed a script for downstream automation
Survey / GNSS coordinates (optional)
- Enter the lip’s latitude, longitude, elevation, and feature heading. Pick WGS84 lat/lon, UTM (auto zone), or any EPSG code — state plane, NAD83, or your local grid
- CSV gains world-coordinate columns in the CRS’s native units (degrees, meters, or US feet) so a GNSS controller can navigate to every stake directly — no manual translation from lip-relative offsets
- DXF gains a WORLD layer with georeferenced 3D geometry — drops into Civil 3D, Trimble Business Center, or QGIS already in the right place
- PDF prints a one-line geo caption (lat/lon, heading, datum, CRS) so the paper plan in the field still ties to the surveyed location
- Leave the toggle off and exports stay lip-relative — the tape-and-stringline workflow is unchanged. Same tool, three crews
In every export
- Cut/fill dirt estimates, gap pit, and net import/surplus when approach and landing-out grades are provided
- Speed handoff table for chained feature sequences
- Full run audit — every result traceable to its inputs
Explicit in.
Explicit out.
A model is only useful if it’s honest about what it doesn’t model. BASE-B’s exclusions are documented by design, not buried. The User Guide is loaded with how-to’s and what the inputs and outputs do, as well educational tips, it’s a masterclass in itself.
- Ballistic point-mass trajectory under gravity, and drag
- Clothoid and clothoid-flat ramp geometries based on G-force limits
- Constant-EFH and Variable landing surfaces by ordinary differential equations (ODE)
- Non-EFH tangent landing option
- Table-top style G-force evaluation option
- G-force options for ramp transitions and landings
- Four-speed flight envelopes (slow / design / fast / pop)
- Site elevation air density correction for aerodynamic drag
- Wheelbase contact geometry at touchdown
- Run-in and speed estimation with friction coefficient options
- Feature chaining with speed hand-off, and friction coefficient and drag throughout
- Cut/fill dirt budget from site grades, feature choices, soil:rock ratios, and excavation decisions
- Borrow pit and quarry options to help dirt budget needs for feature choices
- Assessment mode — evaluate existing features
- Rider input — pedaling or braking
- Rider input — tricks (assume less range.)
- Suspension compliance and rebound dynamics
- Soil conditions, compaction, or material properties
- Wind — envelope visualization is diagnostic only, not a surface design driver
- Tire PSI, behavior or traction
- Environmental variability — moisture, temperature
- Structural integrity of built features
- Turn or berm geometry (planned for v2)
BASE-B produces physics simulation estimates based on idealized mathematical models and simplified assumptions. These outputs are planning aids, not guarantees of safety, performance, or suitability for construction. Any feature informed by BASE-B must be reviewed, approved, and supervised by a qualified professional — licensed engineer, certified bike park designer, or equivalent — before construction and before use by any rider.
Already built something?
Assess it.
BASE-B isn’t only for new construction. Assessment mode runs the full physics pipeline on features that already exist — giving you EFH, touchdown analysis, and trajectory overlays for any feature you create and measure — jumps, drops, and roller chains (including pump tracks with mixed-bump-size geometry) — so you can keep doing what you do as you like to do, and assess whether you want to keep it as is or change it. The goal of BASE-Bf isn’t to be the design police, but to help make informed designs and keep the freedom you currently practice.
Enter the measured key dimensions for your feature — jumps (ramp, lip angle, knuckle, landing), drops (face height + landing), or rollers (height, wavelength, count). BASE-Bf reconstructs a representative profile and runs the full pipeline — EFH grade, three-speed trajectories, touchdown analysis, and the same DXF/PDF exports as a generated design.
Paste x,y points or upload a CSV — works with RTK rover exports after a column trim. AutoQC flags gaps, spikes, and sparse sampling before any physics runs. For pump tracks and roller chains, every crest is auto-detected and analysed individually — no need to identify launch points.
Chain up to five
features in sequence.
Design a full jump line, not just a single feature. Each feature in a chain passes its exit speed into the next as entry speed — automatically. Incline and decline connectors between features let you model the speed change from elevation transitions. Sorry, just 5…For 5 links there are thousands of calculations and plot points being created (each jump feature is ~4500+ ODE evaluations. With 4 features ~18,000 ODE solves. Jumps are the bottleneck because they have both flight + backward-EFH.
Feature 2 and beyond get their entry speed from the computed exit speed of the feature before them. No manual re-entry. The chain result includes a full handoff table showing entry and exit speed at every feature.
Incline/decline connectors between features compute speed gain or loss from slope and friction. Not a standalone feature — chain-only. Lets you model a descent between a roller and a jump, or a short climb between two drops. This is a conservation of energy simulation, thing Trailism wanted from day one, and it is now finally here.
The chain hero view renders all features in global coordinates on a single axis — feature labels, speed annotations at each handoff, three trajectory overlays per feature, and touchdown wheelbase circles at every landing.
v1 — Local desktop/laptop.
v2 — Tablet, Mobile.
BASE-B v1 runs locally on your machine. No account, no cloud, no data sent anywhere. Your designs stay on your computer.
Runs on Windows, Mac, and Linux. Full feature set: jumps, drops, rollers, chaining, assessment mode, DXF + PDF export.
Minimum: Dual-core 2 GHz (2015+ laptop) | 4 GB | 1–3 sec per feature
Recommended: Quad-core 2.5+ GHz (2018+ laptop) | 8 GB | Sub-second, smooth
Ideal: 6-core 3+ GHz | 16 GB | Instant
Tablet-first interface. Visual chain builder. Full module library. iOS and Android. Same backend as v1.
Contact TrailACT for access to the v1 desktop application, or sign up to be notified when the web app launches.
BASE-B v1 runs entirely on your local device. No user data, inputs, outputs, or usage information is transmitted to any server or third party. Your designs and inputs never leave your computer.
Recommended Survey Gear
BASE-BF imports CSV and DXF from any standard surveying tool. These are the kits we recommend by use case — picked for accuracy, build quality, and BASE-BF integration. We don’t sell or affiliate any of these.
| Use case | Kit | Approximate cost |
|---|---|---|
| Weekend builder, open-sky site | Emlid Reach RX2 + NTRIP subscription + Leica Disto D2 | ~$2,200 |
| Professional trail crew, mixed site | Emlid Reach RS4 (base) + RX2 (rover) + Leica Disto S910 | ~$5,700 |
| Pro crew wanting canopy coverage | Emlid Reach RS4 Pro (AR cameras for GNSS-denied spots) + RX2 | ~$6,700 |
| Parks / agency / consultant, mostly wooded | Topcon GT-503 or Leica TS13 (robotic total station) + Emlid RX2 for open areas | ~$22,000–25,000 |
Notes
- Open sky vs canopy decides almost everything. RTK GNSS is fast and one-person, but it loses fix under tree cover. Total stations don’t care about sky but cost 4–10× more.
- Always carry a Leica Disto S910. Handheld 3D laser, ~$1,500. Fastest tool for spot checks, as-built QA, and stakeout verification — pays for itself on small jobs.
- Skip IMU-only tools (Moasure and similar) for trail work. Drift compounds over distance; fine for a residential lawn, not fine for a 40 m feature spine.
- All recommendations export CSV or DXF that BASE-BF imports directly. Drone-photogrammetry workflows (DJI Mavic 3E + RTK) are supported via DEM, but spine extraction is currently manual.
Prices are ballpark and shift with the surveying market — verify at purchase.