Skip to content

Livestock 3D · Phenotyping camera

Transform Your Farm with AI-Powered 3D Monitoring

Revolutionary biometric technology that monitors cattle health 24/7, detects lameness early, and optimizes production through advanced 3D imaging and AI analytics.

Request Demo
Explore Products

Real capture · illustrative values

What the camera reads

Select one to see where the camera reads it on the scan.

The network today

of 3D body scans analysed
Millions
each cow scanned as she walks the lane
Every pass
dairies across the United States and Mexico
US & MX
monitoring, without extra labour
24/7

Live on commercial dairies with the Delta platform, 2026

01 · How it works

How L3D Technology Works

  1. 01

    3D Scanning

    Advanced cameras capture 30 million 3D points per second as cattle walk underneath

  2. 02

    AI Analysis

    Machine learning algorithms analyze body shape, condition, and movement patterns

  3. 03

    Real-time Insights

    Get instant alerts and actionable insights through web and mobile applications

  4. 04

    Secure Data

    Encrypted in transit, role-based access for every user, and your herd data stays yours

A Livestock 3D camera on its steel bracket above a parlour exit lane, seen from below, lenses facing down.
The camera above the lane. One camera hangs over the single-file lane and sees every cow from above as she walks underneath. The livestream below is that view.

See L3D Technology in Action

Watch real-time analysis as cattle walk through the scanning area

Recorded from the Delta livestream on a commercial dairy: the colour view and the depth map of the same moment, where the cow's back is nearest (cyan and green) and the floor falls away to blue. Each cow is outlined as she walks in, turns green once she is measured, and is paired with her ear-tag read.

02 · The camera

Our Products

Core Product

L3D Phenotyping Camera

L3D Phenotyping Camera

  1. 01

    One sealed unit

    Depth sensor, colour camera, thermal sensor and edge computer in an IP68 housing, mounted above the lane.

  2. 02

    60° field of view

    Every cow that walks through the individual exit area is captured from above at each pass.

  3. 03

    Processed on board

    Neural processing on the camera turns each pass into measurements at the lane; results and captures then sync to the Delta platform.

Key Features

  • Body Condition Score (BCS) monitoring
  • Weight estimation without scales
  • Early lameness detection
  • Linear trait measurements
  • Thermal imaging capabilities
  • IP68 protection rating

Technical Specs

Processor
Intel Atom x7-E3950
Neural Processing
3x Intel VPU
Field of View
60°
Connectivity
WiFi, Ethernet, Bluetooth
Capture rate
30 million 3D points / s
Protection
IP68
Request Demo

03 · What it measures

Precise Linear Trait Measurements

Advanced 3D imaging captures detailed body measurements for optimal breeding decisions

The gold overlays show how each trait is read. Values come from one real capture and are illustrative.

WHF — World Holstein-Friesian Federation linear classification

01This capture1.32m

Stature

Measured from top of the spine in between hips to ground.

  • Precise measurement in centimetres or inches, or linear scale.
  • Reference scale: 1.30 m – 1.54 m; 3 cm per point

Linear scale

  1. 1
  2. 2
  3. 3
  4. 4
  5. 5
  6. 6
  7. 7
  8. 8
  9. 9
1
Short1.30 m
5
Intermediate1.42 m
9
Tall1.54 m

A vertical calliper runs from the top of the spine between the hips down to the floor.

02This capture39°

Angularity

The angle and openness of the ribs, combined with flatness of bone avoiding coarseness.

  • Reference scale: weighting of the three components; angle and open rib 80%, bone quality 20%

Linear scale

  1. 1
  2. 2
  3. 3
  4. 4
  5. 5
  6. 6
  7. 7
  8. 8
  9. 9
1–3
Lacks angularity, close ribs, coarse bone
4–6
Intermediate angle with open rib and intermediate bone quality
7–9
Very angular, open ribbed, flat bone

A plumb line and a rib line leave the backline at the withers; the arc between them marks the rib angle.

03This capture0.48m

Chest Width

Camera: Measured as the real width of the chest

  • WHF: Measured from the inside surface between the top of the front legs.

Linear scale

  1. 1
  2. 2
  3. 3
  4. 4
  5. 5
  6. 6
  7. 7
  8. 8
  9. 9
1–3
Narrow
4–6
Intermediate
7–9
Wide

In the top view a dimension line spans the chest from side to side, just behind the front legs.

04This capture0.69m

Body Depth

Distance between the top of spine and bottom of barrel at last rib – the deepest point. Independent of stature.

  • WHF reference scale: optical in relation with the balance of the animal.
  • Camera reference scale: absolute referred to the population.

Linear scale

  1. 1
  2. 2
  3. 3
  4. 4
  5. 5
  6. 6
  7. 7
  8. 8
  9. 9
1–3
Shallow
4–6
Intermediate
7–9
Deep

A dimension line runs from the top of the spine to the bottom of the barrel, just behind the front legs.

04 · The scan

Experience Our 3D Scanning Technology

Interact with a real 3D scan of a cow captured by one of our L3D camera systems

Point cloud of a cow's back from the overhead camera in its real coat colour, seen from above and behind, with gold measurement overlays

Overhead capture, Sep 2026

Overlays as Delta exported them for this capture.

05 · Delta platform

Powerful Analytics Dashboard

Access your herd data anywhere, anytime with our intuitive multi-platform software

Herd at a glance — 2026-09-27

3,412active

4,730 total in herd

  • Milking

    2,860

    avg DIM 171 days

  • Avg BCS

    3.12

    61 below · 408 above

  • Avg locomotion

    1.46

    score 1–5 · lower is better

  • Avg weight

    1,418lb

    mostly estimated

Delta Overview, recreated with anonymised demo data.
Easy AccessWeb, tablet, iPhone

Our web based app allows you to access your herd data easily from wherever you are! Simply log in from your computer, tablet, or phone and get real-time data on each cow!

Real-time Analytics
Monitor herd health with dynamic charts and KPIs
Multi-Platform
Access from desktop, tablet, or smartphone
Smart Alerts
Get notified of health issues immediately
Historical Data
Track trends and patterns over time

Delta on the web and on iPhone — the same dashboards in your pocket, with recent events cached for the barn.

Delta

Delta is where every scan becomes a decision: a herd dashboard, a card for every cow, a live view of the lane and an assistant that answers in plain language.

Locomotion · worklist

42 of 2,860 milking cows · 14 days

Cows to check

The worst walkers the cameras saw, worst first, plus any cow whose locomotion is rising — and what the trimmer actually found on each one.

Herd dashboard

Body condition, locomotion, weight and milk against days in milk, with the share of the herd inside the 2.5–3.5 BCS target band.

Health score and forecast

A 0–100 health score blending milk, body condition and locomotion, anomaly detection, and a 30-day milk projection per cow.

Livestream

Watch the lane live: the colour camera beside the false-colour depth map, with tag reads as cows pass.

A card for every cow

Camera scans against clinical events on one timeline, every scan with its photo, its 3D model and its measurements.

Reports, written and spoken

Scheduled PDF and audio herd reports emailed to your team: what changed, what needs attention, what to do next.

NaDIA · Herd assistant

Ask your herd in plain language.

NaDIA reads your farm's live records — milk, body condition, mobility, weight, health risk — and answers with tables and the exact query behind every number. She replies in your language and only sees what your role allows.

Example conversations with NaDIA

Question: Which cows lost the most condition in the last 30 days?
Answer: 7 cows show a sustained BCS loss of 0.3 or more over 30 days, mostly in early lactation (DIM < 60).
CowDIMBCS 7-dayChange 30d
2417342.71−0.42
3086512.64−0.38
1754222.80−0.35
Question: How many cows are lame this week, by pen?
Answer: Pen 4 has 5 cows at or above the lame threshold on their 7-day average; pens 1 and 2 have 2 each.
PenCowsLameRising
Pen 141224
Pen 238823
Pen 443156
Question: ¿Qué vacas tienen el peso en descenso?
Answer: 9 vacas pierden peso de forma sostenida en los últimos 14 días; 6 están en los primeros 60 días de lactancia.
VacaDIMPeso (lb)Cambio 14d
2960451,318−41
2233581,402−37
3310271,276−33

Works with data from the systems dairies run

Milk, events and roster data flow in from your parlour and herd software, so every scan sits next to the rest of the cow's story.

  • DeLaval DelPro
  • BouMatic
  • GEA
  • SCR / SenseHub
  • DairyComp 305
  • BoviSync

Works with data from these systems through customer-authorized access. Trademarks belong to their owners.

06 · Visual ID · New

The right cow. Every scan.

Visual ID recognises each cow from the same 3D scan and photo our camera already takes. Every body condition score, weight and locomotion reading lands on the right animal, even when a tag read is missed or pinned on the cow behind her.

Live in the Delta platform on US commercial dairies since September 2026.

No new hardware. It reads the capture the camera already makes for body condition, weight and locomotion.

99%+

accuracy

Measured on US commercial dairies, on cows the models never trained on.

Works on dairies it has never seen
It names the right cow from a single scan, even on a farm the model was never trained on.
Proven on commercial herds
Evaluated on cows the models never trained on, in herds of thousands of animals.

Same cow, different days

Top-down frames from the camera, weeks apart and across two lanes. Her markings are her fingerprint.

Six top-down camera frames of the same black-and-white Holstein, taken between July and September. Her markings match in every frame.

  1. Jul

  2. Jul

  3. Aug

  4. Aug

  5. Aug

  6. Sep

How Visual ID works

Tags name the cow. The camera checks every scan.

  1. 01

    Capture

    She walks under the camera. It records a 3D model of her back and a top-down photo — the same capture used for body condition, weight and locomotion.

  2. 02

    Fingerprint

    AI turns that capture into a compact fingerprint of her coat pattern and body shape. It works on patterned and solid-coloured herds.

  3. 03

    Match

    The fingerprint is compared with the cows on your farm — never anyone else's. It commits only when the best match clearly beats the runner-up; otherwise it marks the scan uncertain instead of guessing.

  4. 04

    Confirm

    Where a tag was read, Visual ID confirms it, or flags the mismatch and names the cow it actually saw. Every correction is logged and can be undone.

New cows join from the identity you already have: a tag read or your herd records. A cow it can't place becomes a task for your team, not a guess.

No two coats are alike

Solid coats too. It also works on solid-coloured herds, not only patterned ones.

Wrong cow? Visual ID moves the scan.

When a scan lands on the wrong cow, Visual ID moves it to the cow it actually saw and says why. Every move is logged and can be undone.

Animal detail · Observations timeline

Cow A

  1. Today

    • 05:52BCS 3.251,482 lb
    • 14:37BCS 2.751,318 lb
Values and cow names are invented.
See Visual ID on your herd

07 · Why L3D

Why Choose L3D?

Two Holstein cows stand at a feed rail, bodies side-on to the camera.
  1. 01

    Economic Impact

    A lame cow can cost $90–300 a year. Early lameness detection and optimized nutrition management protect that margin across the herd.

  2. 02

    24/7 Monitoring

    Continuous health tracking without human intervention, providing multiple measurements daily

  3. 03

    Objective Data

    Remove human bias with automated, accurate measurements backed by scientific validation

  1. 04

    Sustainability

    Optimize feed efficiency and reduce environmental impact through precision monitoring

  2. 05

    Animal Welfare

    Non-intrusive monitoring reduces stress while improving early disease detection

  3. 06

    Easy Integration

    Compatible with existing farm infrastructure and management systems

08 · Evidence

Scientific Validation

Our technology is backed by extensive research and university studies

  • University Partnerships

    Evaluated by Ghent University, Teagasc and Institut de l'Élevage across our technology's lineage

  • Published Studies

    A peer-reviewed study and congress presentations demonstrating the accuracy and effectiveness of 3D imaging

  • Field Testing

    Proven results from installations across Europe and North America

Our technology lineage

Studies of the predecessor 3D devices our current camera descends from (2014–2018).

  • r = 0.82

    BCS correlation with an experienced veterinarian; 90% of scores within 0.5 BCS

    Ghent University · Vlaams Diergeneeskundig Tijdschrift, 2016 · peer-reviewed

  • CCC 0.70

    Automated BCS against two expert scorers, who agreed with each other at 0.69

    Teagasc Moorepark · World Buiatrics Congress, Sapporo 2018

  • R² = 0.91

    Live weight of calves against the scale, in a programme covering more than 10,000 calves

    Institut de l'Élevage / FCEL · International Limousin Congress, 2016

University Study Highlights

  • The advantage of BCS automation is that it may provide objective, frequent, and accurate BCS with a higher degree of sensitivity compared with more sporadic and subjective manual BCS.

    Source: Truman et al., 2022[1], see reference

  • Cows that lost BCS during the dry period had increased odds of developing sub-clinical ketosis, metritis, and transition diseases.

    Source: Daros et al., 2020[4], see reference

  • Studies found that the lameness of dairy cows can be correctly detected through analysis of the curvature features of dairy cows' backs.

    Source: Jiang et al., 2022[3], see reference

  • Using 3D imaging, a recent study tracked cows' body growth across lactations. Growth changes dominated most of the body weight gain, stressing ongoing cow growth even into their third lactation.

    Source: Xavier et al., 2022[5], see reference

References5
  1. [1]

    Truman, C. M., Campler, M. R., & Costa, J. H. C. (2022). Body Condition Score Change throughout Lactation Utilizing an Automated BCS System: A Descriptive Study. Animals, 12(5), 601. doi.org/10.3390/ani12050601 (opens in a new tab)

  2. [2]

    Ruchay, A., Kober, V., Dorofeev, K., Kolpakov, V., & Miroshnikov, S. (2020). Accurate body measurement of live cattle using three depth cameras and non-rigid 3-D shape recovery. Computers and Electronics in Agriculture, 179, 105821. doi.org/10.1016/j.compag.2020.105821 (opens in a new tab)

  3. [3]

    Jiang, B., Song, H., Wang, H., & Li, C. (2022). Dairy cow lameness detection using a back curvature feature. Computers and Electronics in Agriculture, 194, 106729. doi.org/10.1016/j.compag.2022.106729 (opens in a new tab)

  4. [4]

    Daros, R. R., Eriksson, H. K., Weary, D. M., & von Keyserlingk, M. A. G. (2020). The relationship between transition period diseases and lameness, feeding time, and body condition during the dry period. Journal of Dairy Science, 103(1), 649–665. doi.org/10.3168/jds.2019-16975 (opens in a new tab)

  5. [5]

    Xavier, C., Le Cozler, Y., Depuille, L., Caillot, A., Lebreton, A., Allain, C., Delouard, J. M., Delattre, L., Luginbühl, T., Faverdin, P., & Fischer, A. (2022). The use of 3-dimensional imaging of Holstein cows to estimate body weight and monitor the composition of body weight change throughout lactation. Journal of Dairy Science, 105(5), 4508–4519. doi.org/10.3168/jds.2021-21337 (opens in a new tab)

Expert Opinion

Cattle metabolism responds to fluctuations in weight and condition rather than static values.

Nutritionists need to make timely adjustments based on these changes to enhance decision-making impact. Utilizing 3D cameras enables the quantification of these variations, facilitating necessary dietary and management strategies modifications to achieve desired animal production and well-being goals.

Dr. Alvaro Garcia - Livestock Dairy NutritionistDr. Alvaro GarciaPhD | Livestock Dairy Nutritionist

10 · Installation

Easy Installation

The installation of the Livestock 3D camera system is quick and easy with only a few simple requirements. Our team is happy to assist in finding the perfect location for the system and help with installation.

System Requirements

In order to ensure accurate data collection of your herd, the following items are required:

  1. 01

    High Speed Internet

    Reliable internet connection for real-time data processing and cloud synchronization

  2. 02

    Solid Mounting Surface

    Flat, stable surface for secure camera mounting at optimal height

  3. 03

    Individual Exit Area

    Space where cows can walk through individually for accurate scanning

Typical placement

Side elevation

Holstein≈ 3.05 m · 10 ftJersey≈ 2.8 m · 9 ft60° field of viewRFID antenna

End view

Single-file lane≤ 80 cm · 32 in
Camera centred over the lane, as close as practical to the ID antenna. Heights run from the floor to the camera and are confirmed on site.
L3D cameras on the steel frame above a parlour exit lane, with a cow looking out from the lane.
L3D Cameras mounted above the parlour exit lane
Two L3D cameras on the steel frames above the entry to a rotary milking parlour, one on the far beam and one on the near beam.
Multiple L3D Cameras installed in milking parlor

11 · Contact

Ready to Transform Your Farm?

Join progressive farmers worldwide who are using L3D technology to improve animal welfare and farm profitability

Request Your Demo

Get in Touch

Optional
Optional
Optional

We use your details to answer your request. Privacy Policy

Contact Information

Swiss Headquarters

Livestock 3D SA

Via Pestariso 5CH-6982 Agno, Switzerland

+41 91 224 18 12

USA Office

Livestock 3D Inc.

161 C John Jefferson RdWilliamsburg, VA 23185, USA

+1 757-758-2239