top of page

Are you ready to have geo-reference data for your business?


We aim at developing a new indoor databank to support intelligent decision making of a mobile robot. Such intelligent decisions may include autonomous navigation, security surveillance, customer flow analysis, customer guidance, and facility management. The idea is that whenever there is a decision related to the real-world environment to be made, the robot needs a library or databank to refer to make a meaningful choice and to perform appropriately.


The problem is that at this moment, there is no such indoor databank in existence. In the outdoor world, overhead satellites, CCTV networks on roads, and cameras and radars in vehicles are supplying sufficient data for rebuilding the real world in digital form.


We adopt a visual approach to record the most meaningful and useful information. We develop our visual machine learning (VML) algorithms and train the AI with our 360° images that are annotated in 3D GIS-based building models. Now, the VML can be specialised in fire management, customer flow analysis, and autonomous navigation.

New framework supporting robotic intelligence based on computer vision, magnetic field and wi-fi fingerprinting technologies

Our autonomous robotic navigation is innovative in that it relies on hybrid magnetic field, wi-fi fingerprinting, visual positioning, and motion sensor fusion navigation technology. Our prototype has been successfully implemented in a commercial robot.  This new system requires only a lightweight camera and moderate computational power. The advantage of the system is that the visual data acquired by the VML not only supports navigation but other intelligent decisions along with the benefit that the associated supporting modulus is simpler in terms of data processing power, the battery system, and the motor systems required.


Implementation of semantic navigation in a commercial mobile robot

Our unique low-cost GeoMove Trackers are integrated with fall detection and indoor-outdoor positioning algorithms to answer the fundamental question "where does the fall happen". The core values of the algorithms are that they enable real-time detection of falls both indoors and outdoors with an affordable wearable smart device. This is the first smart device (other than smartphones) with such a large indoor location tracking capacity and motion sensing ability at such low cost.


We develop our technologies based on magnetic fields and Wi-Fi signal fingerprints, which require no extra infrastructure set up at the sites. This type of positioning technology has a nominal accuracy of 3-5 meters. Our algorithms rely on magnetic fields and Wi-Fi signals as the input data. To have the positioning services, magnetic field and radio frequency survey maps must be built beforehand based on the pre-existing router network. In the case that there is no Wi-Fi network, or the signal coverage is not adequate such as in those construction and mining sites, we also provide industrial-grade routers and Wi-Fi hotspot devices.

Picture 3.png

Location AI hybrid magnetic field and wi-fi fingerprinting positioning system

The smart device is equipped with motion sensors, edge computation modulus (MCU), seamless positioning modulus (GPS and Wi-Fi sensor fusion algorithms), and 4G/5G communication modulus.

Location Ai hotspot.png

Location AI Wi-Fi hotspot

bottom of page