Marvelmind Presentation

The Marvelmind presentation as HTML: problems solved, real use cases, architectures and products.

Jump to a solution: robots / AGVs · drones · DJI · forklifts · people · architectures · costs · all solutions.

Overview

Indoor Positioning System – ±2 cm

Marvelmind is an off-the-shelf indoor positioning system (indoor “GPS” / RTLS) with ±2 cm accuracy for autonomous vehicles, robots, drones, forklifts and people. Stationary ultrasonic beacons are united by radio in the license-free ISM band; a mobile beacon on the tracked object reports its precise location in real time – no GPS and no internet required, working in the GPS-denied, RF-noisy industrial environments where other systems fail.

This page mirrors the Marvelmind presentation as HTML; each topic links to its full solution page. New to setup? See the Placement Manual, the architectures comparison and how it works.

Indoor Positioning System for vehicles, robots, drones, forklifts and people - Marvelmind presentation, slide 1
Slide 1 – Indoor Positioning System for vehicles, robots, drones, forklifts and people · view in PDF
Problem and solution

Why GPS fails indoors

GPS does not work indoors: no direct view of satellites, and where a weak signal leaks in, accuracy is metres, not the centimetres indoor automation needs. Wi-Fi and Bluetooth RSSI give 1-10 m; UWB typically 10-30 cm.

Indoor robots, drones and forklifts need reliable ±2 cm, everywhere, all the time. For the full method-by-method comparison watch the technologies review video and read the technologies review (PDF).

GPS fails indoors - accuracy in metres, not centimetres - Marvelmind presentation, slide 3
Slide 3 – GPS fails indoors – accuracy in metres, not centimetres · view in PDF

The Marvelmind solution

An off-the-shelf ultrasonic system: stationary beacons on walls or ceiling, a mobile beacon on the object, and a modem/router to the PC, ROS or robot. Ultrasound time-of-flight plus multilateration computes the position to ±2 cm in real time.

Submaps tile to cover any size facility. See how the system works and the Placement Manual for where to put the beacons.

The Marvelmind ultrasonic indoor positioning solution - Marvelmind presentation, slide 4
Slide 4 – The Marvelmind ultrasonic indoor positioning solution · view in PDF

Starter Set – ready to use

A typical Starter Set contains 1 mobile beacon, 4 stationary beacons and 1 modem/router, with all software included – ready to evaluate ±2 cm tracking out of the box and expandable with submaps to 200,000 m² and beyond.

For current prices see costs and the shop; browse the full range on the products page.

Starter Set configuration - ready to use, expandable with submaps - Marvelmind presentation, slide 5
Slide 5 – Starter Set configuration – ready to use, expandable with submaps · view in PDF
Proven

On the market since 2014 · 50+ countries

Marvelmind is an autonomous-robotics company headquartered in Tallinn, Estonia, with R&D and support across several countries and production in Shenzhen, China. The indoor positioning system has been on the market since 2014, with customers in 50+ countries across robotics, logistics, industry, safety and research.

Most large and commercial deployments are under NDA; universities and researchers are among the most active public users. See customers and case studies and more about Marvelmind.

On the market since 2014 - customers in 50+ countries - Marvelmind presentation, slide 6
Slide 6 – On the market since 2014 – customers in 50+ countries · view in PDF
Use cases

Autonomous robots and AGVs

Full solution first: robots and AGVs.

Autonomous mobile robots (AMRs) and AGVs need indoor positioning GPS cannot provide. Marvelmind ultrasound RTLS delivers ±2 cm, about 10-15× more precise than UWB (10-30 cm) and ~50× more precise than Bluetooth. It outputs XYZ over USB, UART or CAN, with native ROS / ROS2 drivers and the NMEA 0183 GPS protocol for immediate integration into any navigation stack.

Use case - autonomous delivery robots in warehouses and assembly plants - Marvelmind presentation, slide 7
Slide 7 – Use case – autonomous delivery robots in warehouses and assembly plants · view in PDF

Indoor drones (incl. DJI)

Full pages first: indoor drones, DJI drones, the DJI vs Pixhawk comparison and the Pixhawk integration manual (PDF) from downloads; small craft: positioning for small drones.

Marvelmind gives autonomous indoor drones ±2 cm positioning for inventory, inspection and scanning, and works outdoors too. Integrate via Pixhawk/ArduPilot/PX4 (UBX) or, for DJI drones (from DJI Mini 3), via the DJI SDK and the Marvelmind app: fly waypoints in the Dashboard, take pictures, scan QR/bar codes, stream coordinates to WMS/ERP and return to base. Use NIA – see the architectures comparison for why.

Use case - autonomous indoor drones and DJI drones - Marvelmind presentation, slide 8
Slide 8 – Use case – autonomous indoor drones and DJI drones · view in PDF
Use case - autonomous indoor drones and DJI drones - Marvelmind presentation, slide 13
Slide 13 – Use case – autonomous indoor drones and DJI drones · view in PDF

Forklifts, AGVs and vehicles

Full solution first: forklifts, carts, AGVs and vehicles.

Track them to the centimetre for productivity, safety and traffic analytics: real-time position, full path logged to CSV, geofencing and alarms, and two-way communication with drivers. Mount the mobile beacon on the vehicle; ceiling stationary beacons cover huge open warehouses via submaps.

Use case - tracking forklifts, AGVs and huge transport vehicles - Marvelmind presentation, slide 9
Slide 9 – Use case – tracking forklifts, AGVs and huge transport vehicles · view in PDF
Use case - tracking forklifts, AGVs and huge transport vehicles - Marvelmind presentation, slide 14
Slide 14 – Use case – tracking forklifts, AGVs and huge transport vehicles · view in PDF

People and worker safety

Full page first: people tracking (and the people-tracking video).

Indoor positioning for people raises productivity and safety: locating workers in factories, underground, in metro and tunnels, on construction sites and at railway stations, with reporting and geofencing alarms (visual, vibration, sound) on approach to dangerous zones. It matters most where moving vehicles, falls or restricted areas are a risk and GPS cannot reach. For people choose the Inverse Architecture (IA) so the worn beacon stays silent – see the architectures comparison.

Use case - tracking and protecting people and machinery - Marvelmind presentation, slide 10
Slide 10 – Use case – tracking and protecting people and machinery · view in PDF
Use case - tracking and protecting people and machinery - Marvelmind presentation, slide 12
Slide 12 – Use case – tracking and protecting people and machinery · view in PDF
Use case - tracking and protecting people and machinery - Marvelmind presentation, slide 15
Slide 15 – Use case – tracking and protecting people and machinery · view in PDF
Use case - tracking and protecting people and machinery - Marvelmind presentation, slide 17
Slide 17 – Use case – tracking and protecting people and machinery · view in PDF

Geofencing and 1D distance

Define service and geofencing zones of any size; a zone violation triggers visual, vibration or sound alarms and is logged for analysis – used for construction and crane safety and for AGV-vs-human collision avoidance.

Marvelmind also measures cm-level 1D distance without a laser, where lasers fail (fog, smog, bright light, foliage). With the Marvelmind Horn the 1D range reaches up to ~100 m (120-150 m in the lab) – see what is the range and the range of the indoor positioning system (the 30 m is only the per-submap line of sight; total area is unlimited via submaps).

Use case - geofencing zones, alarms and precise 1D distance - Marvelmind presentation, slide 11
Slide 11 – Use case – geofencing zones, alarms and precise 1D distance · view in PDF
Use case - geofencing zones, alarms and precise 1D distance - Marvelmind presentation, slide 16
Slide 16 – Use case – geofencing zones, alarms and precise 1D distance · view in PDF
How it works

Architectures and very large networks

Full page first: the architectures comparison (NIA vs IA vs MF NIA).

The system runs in NIA, IA or MF NIA; the difference is who emits the ultrasound. Submaps tile to cover any size (the 30 m is only the per-submap line of sight; a 200,000 m² warehouse is no problem), and the multi-modem architecture extends to very large linear networks such as tunnels, mines and long warehouses.

How it works - architectures and the multi-modem architecture - Marvelmind presentation, slide 18
Slide 18 – How it works – architectures and the multi-modem architecture · view in PDF
How it works - architectures and the multi-modem architecture - Marvelmind presentation, slide 19
Slide 19 – How it works – architectures and the multi-modem architecture · view in PDF
How it works - architectures and the multi-modem architecture - Marvelmind presentation, slide 20
Slide 20 – How it works – architectures and the multi-modem architecture · view in PDF
Products

The beacon and modem family

Full page first: all products.

A full family covers every case: Super-Beacon (and Super-Beacon-Outdoor), Beacon Mini-RX, Beacon Mini-TX, Industrial Super-Beacon (Metal / Plastic), Industrial-RX, Super-Modem and Modem HW v5.1.

The Marvelmind beacon and modem family - Marvelmind presentation, slide 21
Slide 21 – The Marvelmind beacon and modem family · view in PDF
Why Marvelmind

±2 cm vs 10-30 cm – about 10× more accurate

Against LIDAR, optical flow, structured light, odometry, GPS or magnetic methods, Marvelmind delivers ±2 cm versus a typical 10-30 cm – about 10× more accurate – at a competitive price, and it keeps working in GPS-denied, RF-noisy, metal-heavy industrial environments where UWB and RSSI systems struggle.

Compared with every alternative: UWB (10-30 cm) competes on accuracy, but Marvelmind is ~10× more accurate and less expensive; Bluetooth and Wi-Fi (RSSI, 2-5 m) only partially compete and share the limits of any RSSI technology; LIDAR, inertial, optical flow, laser triangulation, odometry, GPS and magnetic methods complement rather than replace it. For the full, detailed breakdown watch the indoor positioning technologies review video (with transcript) (on YouTube) and read the technologies review (PDF).

Why Marvelmind - ±2 cm vs 10-30 cm, about 10× more accurate - Marvelmind presentation, slide 22
Slide 22 – Why Marvelmind – ±2 cm vs 10-30 cm, about 10× more accurate · view in PDF
Why Marvelmind - ±2 cm vs 10-30 cm, about 10× more accurate - Marvelmind presentation, slide 23
Slide 23 – Why Marvelmind – ±2 cm vs 10-30 cm, about 10× more accurate · view in PDF
Robot v100

Marvelmind Robot v100 – autonomous delivery

Full pages first: the Robot v100 product page and the dedicated Robot v100 website (marvelmind.eu).

The Marvelmind Robot v100 is a fully autonomous delivery robot for smart warehousing and industrial applications – up to 100 kg payload, >16 h drive time, 4,990 EUR. It navigates with the inexpensive Marvelmind indoor “GPS” plus collision-avoidance sensors.

Unlike bulky, expensive, hard-to-integrate traditional AGVs, the v100 is flexible, modular, easy to assemble and maintain, and far more cost-efficient – complementing rather than replacing existing AGV fleets for small-goods delivery in warehouses, retail and assembly plants.

Marvelmind Robot v100 - autonomous delivery robot - Marvelmind presentation, slide 25
Slide 25 – Marvelmind Robot v100 – autonomous delivery robot · view in PDF
Marvelmind Robot v100 - autonomous delivery robot - Marvelmind presentation, slide 26
Slide 26 – Marvelmind Robot v100 – autonomous delivery robot · view in PDF
Marvelmind Robot v100 - autonomous delivery robot - Marvelmind presentation, slide 28
Slide 28 – Marvelmind Robot v100 – autonomous delivery robot · view in PDF
Marvelmind Robot v100 - autonomous delivery robot - Marvelmind presentation, slide 29
Slide 29 – Marvelmind Robot v100 – autonomous delivery robot · view in PDF
Marvelmind Robot v100 - autonomous delivery robot - Marvelmind presentation, slide 30
Slide 30 – Marvelmind Robot v100 – autonomous delivery robot · view in PDF
Marvelmind Robot v100 - autonomous delivery robot - Marvelmind presentation, slide 32
Slide 32 – Marvelmind Robot v100 – autonomous delivery robot · view in PDF
Contacts

Contacts

Next steps: pick your case in all solutions, compare the architectures, check costs and the shop, or ask the AI assistant. New to setup? Start with the Placement Manual.

Marvelmind Robotics · Marvelmind OÜ, Katusepapi tn 4/2, Tallinn 11412, Estonia · info@marvelmind.com · marvelmind.com.

Thank you - Marvelmind Robotics - Marvelmind presentation, slide 33
Slide 33 – Thank you – Marvelmind Robotics · view in PDF
Scroll to Top