Skip to content
IEMI Labs
Now part of RoadOne. RoadOne does this product's jobs and more in one device, our one product for Public sector, roads and airfields.

RoadSense AI

Traffic and active-travel counts as a subscription, per sensor

Target price from$1,490per sensor (waived on a 36-month subscription)

Software: $95–$149 per month per sensor

Concept, not for sale yet. Tell us if you want it; that decides what we build.

Rate this idea and tell us what to change
RoadSense AI concept: the idea and where each part goes

Drag to rotate · right-drag to move · scroll to zoom · click a part for its size and placeSwipe to rotate · pinch to zoom · two fingers to move · tap a part for its size

How it fits together

RoadSense AI, part by part

Scroll or tap a part. Press “See the 3D CAD design” to watch the model, built from our CAD, light each part up and show how it connects.

RoadSense AI concept: the idea and where each part goes

Real-time 3D render · drag to rotate · final design may varySwipe sideways to rotate

  1. 01

    The whole RoadSense AI

    A pole camera records an encrypted traffic study on a memory card; the office hub counts vehicles by turn and class from the video.

  2. 02Brain

    Raspberry Pi Zero 2 W

    The brain: it runs the rs-rec program

  3. 03

    Camera Module 3 Wide

    Films the intersection at 720p, 15 frames a second

    Raspberry Pi Zero 2 W · CSI ribbon cable · Video frames

  4. 04

    DS3231 clock

    Keeps true time so every count lines up with the clock

    Raspberry Pi Zero 2 W · I²C plus 1 Hz GPIO · Time and tick

  5. 05

    256 GB microSD

    Stores the encrypted 5-minute video segments

    Raspberry Pi Zero 2 W · microSD slot · Encrypted video

  6. 06

    5 V step-down

    Turns 12.8 V from the battery into 5 V

    Raspberry Pi Zero 2 W · 5 V power pins · Power

  7. 07

    LiFePO4 battery case

    Powers the camera for a multi-day study from the post foot

    Raspberry Pi Zero 2 W · 2-core cable, IP68 plug · 12.8 V power

  8. 08

    Cat-M status modem

    Optional: sends battery and view status every 15 minutes

    Raspberry Pi Zero 2 W · UART (3 wires) · Status messages

  9. 09

    All connected

    7 key parts, one program, and your phone

What it is

A weatherproof camera sensor on a lamp column that counts road users and sends only numbers.

What it does

  • Counts vehicles, bikes and pedestrians
  • Classifies 8+ road-user types
  • Measures turning movements and speeds
  • Delivers reports for schemes and bids

Who it's for

Councils, transport authorities, DOTs and consultancies monitoring streets.

How the parts connect

RoadSense AI, wired up

Every part, what links it to the brain, and how the data reaches your phone. A solid line is a cable; a dashed line is a radio link.

Camera Module 3 WideVideo framesCSI ribbon cableDS3231 clockTime and tickI²C plus 1 Hz GPIO256 GB microSDEncrypted videomicroSD slot5 V step-downPower5 V power pinsLiFePO4 battery case12.8 V power2-core cable, IP68 plugCat-M status modemStatus messagesUART (3 wires)THE BRAINRaspberry Pi Zero 2 Wruns rs-recNode.js with rpicam-vid and agePhonevia Wi-Fi access point (setup)Edge Hubvia microSD card carried backCloud portalvia MQTT over TLS
Drawn from the RoadSense AI parts list. The 3D view above is built separately, from the CAD.

The program

What controls RoadSense AI

Runs on
Raspberry Pi Zero 2 W
Written in
Node.js with rpicam-vid and age
System
Raspberry Pi OS Lite, read-only root
AI model
None on the pole; the office hub runs YOLOv8s on a Hailo chip
  1. 1Wait for the scheduled start; the RTC sets the alarm
  2. 2Record 720p15 H.264 in 5-minute segments
  3. 3Encrypt each segment with the firm's public key
  4. 4Check every 5 minutes that the view has not moved
  5. 5Stop cleanly when the battery or card runs low
rs-rec.tsmain loop
while (state === 'RECORDING') {
  const seg = await rpicamVid({ width: 1280, height: 720, fps: 15,
                                bitrate: '2M', seconds: 300 });
  await age.encrypt(seg.file, FIRM_PUBLIC_KEY);   // X25519
  fs.rm(seg.file);                                // only the .age file stays
  sidecar.append({ start: seg.start, frames: seg.frames, lux: seg.lux });
  const view = viewMatch(referenceFrame, seg.lastFrame);
  if (view < 0.5) mqtt.event('view_moved', view);
  if (battery.volts() < 12.0) { state = 'LOWBATT'; manifest.end('END_LOWBATT'); }
  if (rtc.now() >= schedule.stop) state = 'SLEEP';
}

To your phone

How the data reaches RoadSense web app

Phone for setup; browser on the office hub

  1. Recorder
  2. Phone
  3. Edge Hub
  4. Cloud portal

What you see in the app

One screen for each job RoadSense AI does.

  • Pole setupA technician straps the recorder to a sign post at a junction. The phone shows the camera view so the whole junction is in frame.

  • Study runningWhile the study runs, the pole sends a short status every 15 minutes: battery, card space and whether the view has moved.

  • Counts readyBack at the office the hub turns the video into classified counts overnight. The firm gets turning counts by class with a quality report.

App screens with sample data. The interface is in development.

What the app gets

Data RoadSense AI sends to the app
Vehicle count18,422 enteringEach job
Count by classLight 17,011; heavy 1,032Each job
Pedestrian count412Each job
Completeness97.3%Each 15 min bin
Pole battery13.12 VEvery 15 min (modem)

Video is encrypted on the pole; the key stays on the hub. No plate or face reading; counts only.

CAD design · Design complete (stages 1-6), not built

Designed in code, checked in 3D

  • 9 printed parts, each a single closed solid
  • Every bought part fitted inside and checked for collisions
  • Assembled exactly as shown in the 3D CAD view (the button on the picture at the top)
See the CAD parts renderRoadSense Count Kit printable parts rendered from the CAD

Key specs

RoadSense AI hardware
AI processorHailo-8L (13 TOPS) + ARM host
CameraGlobal-shutter colour camera (Sony IMX296 class), wide-angle lens with IR-corrected optics
SensorsAmbient light sensor · Tilt / tamper accelerometer · Internal temperature sensor
Connectivity4G LTE (Cat 4) with LTE-M fallback · Optional Ethernet · REST API, CSV and open-data feeds
PowerLamp-column or mains supply (100–240 V AC or 12–24 V DC), ~8 W (target)
EnclosureGrey weatherproof housing with sunshield and stainless pole / lamp-column bracket
Dimensions22 × 14 × 10 cm
Weight~1.4 kg with bracket
Ingress ratingIP66
All target specifications

Detection

Detection target specifications
Classes8+ road-user classes
Count accuracy≥95% for vehicles, ≥90% for cyclists (target)
RangeUp to 4 lanes + footways
NightStreet-lit scenes; IR option planned

AI and data

AI and data target specifications
ProcessorHailo-8L (13 TOPS)
OutputsCounts, turning movements, speeds
IntervalAggregated every 5 min
VideoNot stored or transmitted

Power and physical

Power and physical target specifications
PowerMains / lamp column, ~8 W
Size22 × 14 × 10 cm
RatingIP66
Operating temp.−25 °C to 55 °C

Targets for production units; they may change during development.

Pricing

Hardware: $1,490 per sensor (waived on a 36-month subscription)

Software: $95–$149 per month per sensor

Target pricing. Subscription includes LTE data, dashboard, maintenance and replacement. Installation by a local contractor or our partner network is quoted per site.

Count

$95/mo per sensor

Classified counts for corridors and active-travel schemes.

  • 8+ classes
  • 5-minute counts
  • Dashboard and CSV
  • Maintenance included

Junction

$149/mo per sensor

Adds turning movements and speeds.

  • Turning movements
  • Speed estimation
  • API and open-data feeds
  • Before / after reports

Network

Custom

For authority-wide deployments and consultancies.

  • Volume pricing
  • Integration support
  • Data-completeness SLA
  • Custom reporting

Indicative launch pricing in USD, excluding taxes and shipping. Final pricing is confirmed before any balance is charged.

Community

What do you think of RoadSense AI?

RoadSense AI is in development. Rate the idea, tell us what to change about the features, specs or price, and we'll build it with your feedback.

How much people want it

–

No ratings yet

  1. 50
  2. 40
  3. 30
  4. 20
  5. 10

Follow RoadSense AI

Get build updates and first access when pre-orders open.

Leave feedback

Your rating of the idea

Feedback on a product concept in development, not a review of a shipped product. Posts may be moderated; links and contact details are held for review. See our community guidelines and privacy policy.

Comments

    No comments yet. Tell us what you'd want from RoadSense AI.

    Questions

    Ask about RoadSense AI

    Tell us what this would save you, or ask anything. We reply within two business days.

    We use your details only to respond to this request. See our privacy policy.

    From the product lab · ideation phase

    Lab updates for RoadSense AI

    All lab updates
    1. LatestA council or DOT needs one thing at each signalised crossing that calls the green man for people waiting, holds it for slow walkers, places vehicle demand and counts every road user all year. The MUTCD keeps phases with the controller and lets detectors place calls; video detection needs a camera per approach, often on its own arm (about $15,000 for a four-approach intersection in a Cal Poly study), and street counters go on lamp columns; US DOT Safe Streets and Roads for All grants fund this safety data.
      • Now part of RoadOne: https://iemilabs.com/products/roadone/

      See it in 3DFull research and design

    2. A builders’ merchant, timber yard or garden centre lets customers drive in, park, queue at the will-call counter and walk among forklifts. WorkSafe NZ fined a timber yard after a forklift knocked a pack onto a customer and asked yards to control people on foot and show customers where to park. Today each job is a separate tool: about $950 a truck for an AI reversing camera, $35 a month per camera for plate reading, paid processing for traffic counts and $859-919 per people counter.
      • Combined into SafeYard (https://iemilabs.com/products/safeyard/); RoadSense AI itself is unchanged

      See it in 3DFull research and design

    RoadSense AI FAQ

    General FAQ
    Does RoadSense record video or number plates?

    No. Video is processed on the sensor and discarded. Only counts, classes, turning movements and speeds leave the device. Number-plate reading is not included. A single setup snapshot is used during configuration and can be deleted afterwards.

    How is it installed?

    The sensor is banded to a lamp column or pole at 4–8 m and powered from the column feed or a nearby supply. Installation is usually done by the authority's street-lighting contractor or our partner network and takes under an hour per site.

    What connectivity does it need?

    RoadSense has built-in 4G LTE with LTE-M fallback, and the data plan is included in the subscription. If coverage drops, counts are buffered on the sensor and uploaded later.

    Can we buy through a public-sector framework?

    Not yet. We plan to apply to relevant frameworks such as the UK Crown Commercial Service transport-technology and G-Cloud agreements. Until then, pilots can be bought directly under local purchasing thresholds.

    Is RoadSense AI available now?

    No. RoadSense AI is a new concept in development. Specifications are target specifications. It is a concept: we are not taking orders, reservations or deposits, and it has no ship date yet.

    What warranty will it come with?

    Hardware is covered for the full term of your subscription, including replacement. For customers who buy the hardware outright, we plan a 2-year limited warranty. Final terms will be published before shipping begins.

    Related products

    FlowCount AI concept: the idea and where each part goes
    VisionConcept

    FlowCount AI

    A small white ceiling sensor with two lenses that counts people without sending images.

    • Measures queue length in real time
    • Predicts wait times per lane
    • Counts entries, exits and occupancy

    From $899 per sensor

    Details
    SmokeSentinel AI concept: the idea and where each part goes
    VisionConcept

    SmokeSentinel AI

    A solar-powered four-camera head on a mast that watches the horizon for wildfire smoke.

    • Scans 360° every minute
    • Detects smoke plumes on-device
    • Alerts over LTE or satellite

    From $4,900 per node incl. solar kit

    Details
    AirfieldWatch AI concept: the idea and where each part goes
    VisionConcept

    AirfieldWatch AI

    A thermal and visible camera turret on a mast that spots birds and wildlife near runways.

    • Detects birds and flocks day and night
    • Classifies hazard-relevant species groups
    • Triggers dispersal devices automatically

    From $8,900 per turret

    Details

    RoadSense AI

    Concept · target $1,490

    I want this