Area: Smart Cities · TRL: 3 of 9 (proof of concept on paper) · Value-engineering target: $250 USD for the BridgePulse-specific parts (a hypothetical control target, not a limit) · Difficulty: 3 of 5
A vibration and strain monitor for small bridges and footbridges that tracks natural frequency and strain over time to flag changes that need inspection.
Exploded render · Detail render · Interactive 3D model · Concept blueprint (PDF) · General arrangement BRP-DWG-001 (PDF) · Calculations BRP-CAL-001 · Prototype build plan BRP-BLD-001 · Design decisions BRP-DEC-001 · Review note
A bridge's natural frequencies depend on its stiffness, so a loss of stiffness from corrosion, cracking or a failing bearing shows up as a small, lasting drop in frequency, and a change in how load is carried shows up in strain. The difficulty is that temperature moves the frequencies too. The Z24 bridge study showed that a model of frequency against temperature, learned from the healthy bridge, can separate the two (Peeters and De Roeck, 2001). BridgePulse applies that idea with one clamp-on sensor hub at midspan, hourly 10 minute records and a learned temperature band, so the owner's engineer gets a reason to inspect sooner rather than a stream of raw data.
It is open and garage-buildable because the owners of small bridges, such as towns, counties, park services and rural access programs, are the ones least able to pay for commercial monitoring. Low-noise MEMS accelerometers now cost tens of dollars, the hub clamps on without drilling, and power, radio and data handling come from the lab's shared FieldNode, TwinKit and CityTwin projects. Open design files and open methods also let engineers check how a flag was raised, which matters more than the hardware.
Bridge stock is large, aging and inspected at long intervals. In the United States, 41,685 of 624,193 bridges (6.7 %) were in poor condition in 2025 (FHWA), and routine inspections may be 24, 48 or, under a risk-based method, up to 72 months apart (23 CFR 650.311). In Japan, about 37 % of roughly 730,000 road bridges were 50 years old or more in 2023, rising to about 54 % by 2030 (MLIT).
Findings also have to reach someone who acts. The Fern Hollow Bridge in Pittsburgh collapsed in January 2022 after the city failed to act on repeated maintenance recommendations in its inspection reports (NTSB). A monitor cannot replace inspection or funding, but a clear, early signal on a specific bridge can help put limited inspection time where it is needed.
| Industry | Use |
|---|---|
| Municipal and county public works | Watching footbridges and short road bridges between routine inspections, and after floods or vehicle strikes |
| Parks, trails and recreation | Timber and steel trail bridges that see few inspections and many visitors |
| Rural access and humanitarian programs | Checking a sample of trail bridges built years earlier without a site visit to each one |
| Rail and transit operators | Station footbridges and pedestrian overbridges |
| Forestry and mining | Haul road and forestry bridges carrying heavy, irregular loads |
| Universities and training | An open platform for teaching and research in vibration-based structural health monitoring |
| Country or region | Why it matters there |
|---|---|
| United States | 41,685 bridges were rated poor in 2025 (FHWA); many are small, locally owned structures. |
| Great Britain | 2,928 council-maintained road bridges were substandard in 2023, with a £6.8 billion backlog; councils also report growing concern about flooding and scour (RAC Foundation). |
| Japan | The share of road bridges 50 years old or more is set to rise from about 37 % to about 54 % between 2023 and 2030 (MLIT). |
| Nicaragua and Central America | Floods cut labor income by about 18 % in villages without a footbridge, and new bridges removed that loss (Brooks and Donovan, 2020); keeping existing footbridges in service protects that gain. |
| Rwanda and East Africa | Trail bridge programs report a preliminary annual return of about 49 % on trail bridges in Rwanda and similar settings (Fika, formerly Bridges to Prosperity), so a growing stock of footbridges will need low-cost ways to check them. |
The idea traces back to the Silver Bridge between Point Pleasant, West Virginia, and Ohio, which fell on December 15, 1967, killing 46 people, after a single eyebar in a suspension chain fractured from a defect only 2.5 mm (0.1 in) deep. The disaster led to the National Bridge Inspection Standards of April 27, 1971, which set inspections at least once every two years by trained inspectors (FHWA). That calendar-based regime is still the backbone of bridge safety, but it leaves long gaps in which nobody watches a structure, and it is weakest for the small footbridges and local spans that get the least attention. BridgePulse starts from that gap: a low-cost record of how a small bridge behaves between inspections, fed to the owner's engineer and to CityTwin, so that a change can prompt a visit sooner than the calendar would.
Many small bridges are inspected rarely, and deterioration is found late. Their owners seldom have the budget for commercial monitoring, so there is no record of how a bridge behaves between inspections. See the problem statement.
A vibration and strain monitor for small bridges and footbridges that tracks natural frequency and strain over time to flag changes that need inspection. A sensor hub fixed between the flanges of a girder at midspan records 10 minutes of acceleration, strain and temperature every hour, extracts the natural frequencies and strain statistics on board, and sends about 36 bytes per hour (11 bytes where the radio data rate allows no more) through a FieldNode core over LoRaWAN to TwinKit or CityTwin. A temperature-compensated baseline, learned over weeks, flags lasting changes to the owner's engineer.
The TRL 3 calculations (BRP-CAL-001) give 16.9 mW average draw, 0.0153 Hz spectral resolution per record, 1.39 GB of raw records per month kept on site, and $275 of BridgePulse-specific parts ($401 with the FieldNode core, which is costed in its own repo), $25 over the $250 value-engineering target since the parts that make the design buildable, the lanyard anchor and the installation banding were added. With the lanyard's girder clamp the monitor puts 3.58 kg on the girder. On the 7 m example footbridge the first mode is 23.9 Hz and the instrument can track it to about 0.05 % an hour, but the mass of people crossing lowers it by up to 6 %. The hub therefore gates out the time when someone is on the span; in simulation that brings the hourly scatter to 0.015 % or less, inside the 0.2 % target, which remains at risk until recorded data confirm it. The 1 % change detection cannot be verified on paper. See the requirements for what is and is not met.
Full design precis: docs/02-concept.md
- Sensor hub: IP67 die-cast aluminium enclosure on an aluminium plate that lies on the girder web, clamped to the bottom flange and jacked against the top flange, no drilling
- Low-noise 3-axis MEMS accelerometer (ADXL355 class)
- Signal board with 24-bit bridge ADC, RP2040 class microcontroller, RS-485 link and industrial microSD card
- Two strain gauge half-bridges with temperature-compensating dummy gauges
- Two TMP1826 class temperature probes in stainless sheaths (steel and air)
- Shielded sensor cables with M12 connectors
- FieldNode core (shared lab component): 6 W panel, LiFePO4 cell, MPPT charger and LoRaWAN radio
The working bill of materials is in bom/bom.csv.
The prototype build plan (BRP-BLD-001) shows, component by component and step by step in pictures drawn from the model, how to build the first proof-of-concept monitor on a bench rig of two girder offcuts and a handrail post stub. Seven parts are made in a small workshop from aluminium and steel bar and plate; the rest are bought, and the FieldNode core is built to its own plan. Making the design buildable kept the plate clear of the girder's root fillets, clamped it to the flange with foot blocks and steel jaws, and gave every sensor and cable a fixing that needs no drilling (BRP-DDR-004). A stainless wire lanyard runs from a pad eye on the plate to a second, independent girder clamp on the bottom flange. It is a plan, not yet built; the decisions behind it are in the design decisions register.
A research prototype; it supports but never replaces inspection by qualified engineers, and a "no change" result never means a bridge is safe. Install only with the asset owner's permission, by trained crews, with fall protection, a rescue plan for work over water and traffic management as local rules require. Test old paint for lead before removing any for strain gauges. The FieldNode core contains a lithium iron phosphate cell; use its protected, fused pack. Nothing is drilled or welded into the structure, and the mounting plate carries a secondary wire lanyard to an independent girder clamp.
| Folder | Contents |
|---|---|
docs/ |
Problem, concept, requirements, calculations and design decisions |
cad/src/ |
build123d Python source, the source of truth for all geometry |
cad/step/, cad/stl/ |
Exported models for FreeCAD, other CAD tools and printing |
cad/drawings/ |
2D sketches and dimensioned drawings |
bom/ |
Bill of materials |
electronics/ |
KiCad schematics and PCB layouts |
firmware/ |
Microcontroller code |
media/ |
Renders, perspectives and photos |
build-log/ |
Dated prototyping notes |
Controlled documents follow the portfolio documentation standard. Each carries a document ID (BRP-PRC-001 for the precis), a version and a revision history. Branded PDFs are built with python .kit/render.py and attached to GitHub Releases when a document is tagged, for example BRP-PRC-001/v1.0.
Designed by Amish Chadha. See CONTRIBUTORS.md for roles. To cite this design, use CITATION.cff (GitHub shows it as "Cite this repository").
AI assistance (Claude) was used to accelerate concept renders, prototype documentation and first-pass sizing calculations. Design direction and all decisions are Amish Chadha's, recorded in this repository's decision records (docs/decisions/).
- Hardware (CAD, drawings, BOM, electronics): CERN-OHL-S v2
- Software (firmware, scripts, notebooks): MIT
A project of the Design Molecule lab.