Skip to main content
JANUS AI LINK
中文EN
Contact us

ISAC · Integrated sensing and communication

One radio signal. It communicates, and it senses.

We let a radio signal carry connectivity while measuring and recognising the physical space it passes through. Integration here means sharing the same wireless infrastructure and site assets — not building a second, separate sensing system.

Book a demo See the technology direction
30cm
Indoor 3D positioning accuracy
Decimetre-level — enough for proximity decisions and zone control
0
Cameras needed for sensing
Contact-free sensing captures no images, so there is no privacy trade-off
1
Wireless infrastructure shared
Connectivity and sensing reuse the same site and network assets

What this line does

Not three separate products — three kinds of output from the same wireless infrastructure.

Capability 01

High-precision indoor 3D positioning

Angle-of-arrival measurement resolves the direction of a target relative to each base station; several stations intersect to give a three-dimensional coordinate — which floor someone is on, which side of a machine they are standing on. That third dimension is what industrial sites actually need.

  • Decimetre-level 3D positioning, reaching the 30 cm range
  • Continuous indoor coverage, independent of satellite signals
  • Low-power tags, in badge, helmet or wristband form
Capability 02

Contact-free spatial sensing

Millimetre-wave radar returns tell us whether a space is occupied, where the person is and what state of motion they are in — without capturing a single image. What comes out are structured events, not pictures.

  • Presence, position and activity state recognition
  • No image capture, so no privacy exposure
  • Unaffected by low light, backlight, dust or smoke
Capability 03

Unified deployment

Sensing and connectivity share the same wireless infrastructure and site assets instead of being built as two systems. The practical effect: one less set of equipment, one less installation, one less link to maintain for years.

  • Reuses existing site and network assets
  • Minimal physical modification; runs alongside existing systems
  • Sensing output is consumed directly by business systems

Millimetre-wave indoor sensing demo

No cameras — radar returns tell us whether someone is there, where they are and what they are doing. The panel below is interactive: switch clips and the radar motion energy, chest echo phase, audio energy and fused events update in step.

A multimodal analysis panel: radar, audio and video aligned on one timeline, with the current-moment analysis and fused risk verdict alongside it. Switching sections opens the radar timeline, the audio analysis and the event report.

Open the full demo in a new window

Why "no cameras" comes first

Offices, hospital wards, dormitories, changing areas, plant workfaces — these places are highly sensitive to video capture. Vision is also not technically superior here; it is simply cheaper.

  • Privacy: once video is captured it can be retained, leaked or reused. Contact-free sensing outputs only who is where and what they are doing — never a reconstructable image
  • Occlusion: cameras have blind spots. Radio signals tolerate far more variation in space layout and obstruction
  • Environment: in low light, backlight, at night or in dust and smoke, vision-based systems degrade noticeably
  • Installation: an approach that does not depend on cameras is far less demanding about mounting positions and angles, which simplifies both deployment and recurring maintenance
  • Optional layering: if you already have video surveillance on site, we can reuse those existing points to add behaviour recognition — no new cameras, no extra image retention. But positioning stays with radio, because "who is where, on which floor" is what video struggles to get right

What these capabilities became on industrial sites

Positioning and sensing are only parameters on their own. They become capability once they are wired into business systems. In heavy-asset industrial settings we have turned them into the following.

  • Positioning results mapped onto a 3D model of the plant, showing the spatial relationship between people and equipment — not a dot on a 2D plan
  • Work zones and hazard zones drawn directly in three dimensions, forming geofences that fire automatically
  • Integrated with the permit-to-work system — the permitted area is the geofence, which prevents the classic "the permit ended but the person is still inside"
  • Behaviour recognition and wearable terminals layered on top: helmets, safety harnesses, workwear, plus a one-touch SOS on the tag
  • Full position history retained and replayable, turning "somewhere in that area" into a reviewable time and place

Typical applications

Described as "what the site runs into — how we handle it", so you can map it onto your own situation.

Indoor positioning and navigation

What we see on site

Finding people and assets in large venues, campuses and plants is hard; positioning that gives only a 2D coordinate is often a floor out in multi-storey buildings

How we usually handle it

Replace 2D positioning with 3D, resolving floor and height together; feed results into your existing map or business system to support navigation and dispatch directly

Zone control and unauthorised access

What we see on site

Critical areas need control, but installing cameras triggers employee resistance; access control stops people at the door but not once they are inside

How we usually handle it

Replace video surveillance with contact-free sensing; an unauthorised entry produces an event and an alert, with no image record created at all

Presence and occupancy recognition

What we see on site

Whether a meeting room, desk or common area is occupied — and for how long — can only be checked by hand or with extra sensors

How we usually handle it

Wireless signals determine presence and activity state, outputting occupied / vacant / seated / away states that business systems can consume directly

Industrial site personnel safety

What we see on site

Complex plant layouts and many hazard zones. Watching by eye and assigning blame afterwards never answers "who is inside a hazard zone right now"

How we usually handle it

Real-time control through 3D positioning and geofences; boundary, over-occupancy and proximity alerts fire immediately, and full position history stays replayable

Camera-free retrofits

What we see on site

An existing site needs sensing capability but has no cabling conditions, or will not accept additional camera points

How we usually handle it

Reuse existing site and network assets to add sensing, keeping the scope of physical modification to a minimum

How deployment works

This line needs on-site calibration, so we put validation before investment.

  • STEP 01

    Scenario and site survey

    Establish whether the goal is positioning, zone control or state recognition, and survey site structure, area and existing network conditions.

  • STEP 02

    Point design and simulation

    Produce a point layout and assess accuracy and reachability before any physical work begins.

  • STEP 03

    Deployment and calibration

    Deploy to plan and calibrate on site, delivering a measured accuracy report rather than a theoretical figure.

  • STEP 04

    Integration and iteration

    Connect to your existing systems and validate against agreed criteria; keep calibrating against measured data afterwards.

Frequently asked questions

Under what conditions was the 30 cm figure measured?

It is the indoor positioning accuracy figure we publish for this line. Real performance varies with site structure, area, occlusion and point density. We won't use one number to cover every site — the proper way to judge it is to give us your site details, take our measured expectation for that site, and verify it in a post-deployment report.

How is this different from UWB, Bluetooth or Wi-Fi positioning?

Mainly the order of magnitude and the cost of implementation. Signal-strength-based Bluetooth and Wi-Fi usually land in the metre range; UWB reaches decimetres reliably but needs dedicated tags and hardware, which costs more. Since Bluetooth 5.1 introduced angle-of-arrival measurement, Bluetooth can also reach the decimetre range with noticeably lower tag power and cost. We select the technology to suit the site rather than assuming one — UWB where interference resistance matters most, Bluetooth direction-finding where large numbers of low-power tags matter more.

Does it require modifying the existing network?

It depends on conditions. The principle is to reuse existing site and link assets wherever possible and keep physical modification minimal. The exact scope can only be stated after a site survey — we won't commit to a scope without having seen the site.

What exactly can contact-free sensing recognise?

Currently it addresses presence and state: whether someone is there, whether they are seated, whether they are walking or still, and gesture-level motion. It outputs structured events, not images. For how far it goes in a specific scenario, the demo is far more reliable than any written description.

Can this positioning capability be used for safety management directly?

Yes — and that is one of its most valuable uses. Once 3D positioning, 3D geofences and the permit-to-work system are connected, boundary breaches, over-occupancy, proximity to hazards and unsecured helmets or harnesses all become automatic alerts, while position history turns after-the-fact blame into after-the-fact review. The full solution is on the Products & Solutions page.

Want to see what it can do on your site?

Tell us the floor area, structural characteristics and the problem you need solved. We'll start with an accuracy assessment before deciding whether to move to measured trials.

Book a demo See the industrial safety solution