CNTXT launches robot training facilities in the UAE and expands data collection in Jordan

Intelligent systems are moving away from closed screens to engage with the physical world, but this shift runs into an invisible barrier of scarce fine-grained motion data. CNTXT AI, based in Abu Dhabi, announced the creation of specialised facilities for robot data and training in the United Arab Emirates as part of an investment plan of up to $50 million in robotics hardware, its distribution and operational centres across the United States and the Middle East.

This field expansion is a direct operational step, with the new sites covering 30,000 square feet split between Dubai and Abu Dhabi, and the first facility expected to start operations in November, followed by the remaining sites in turn. According to co-founder and CEO Mohamed Abu Sheikh, these centres will give the country a major strategic advantage in the physical AI sector, alongside accelerating global interest in developing automated machines.

Data collection at the Dubai facility relies on trained workers wearing suits equipped with ultra-sensitive motion sensors to help robots perform the varied tasks normally carried out inside homes or workplaces. Other sites will focus on capturing “egocentric data”, where participants wear lightweight head-mounted gear to record activities from their own viewpoint, allowing precise geometric documentation of hand movements, surrounding context and live tool interaction.

This architecture redefines the notion of digital sovereignty by moving it from text servers and cloud-computing centres to motion sensors in real operational environments.

The company was founded in Abu Dhabi in 2024 as a data-first firm to help organisations build sovereign AI solutions and deploy them while retaining full control over their exclusive data. Its flagship product line includes the “Mansat” model for converting Arabic speech to text, trained on more than 30,000 hours of realistic audio recordings covering over 25 specific Arabic dialects.

These expansions are based on disclosed financial flows; the company has surpassed the $300 million revenue target and earlier this year closed a Series A financing round of $60 million jointly led by Blue Five Capital and AI71. The regional expansion coincides with entry into the North American market through the opening of the first office in San Francisco, California, where most of its clients are currently located, according to its executive leadership.

The field belt also extends beyond the UAE; the system plans to open an egocentric data-collection site in Jordan operating 24 hours a day by the end of November, alongside another facility in India. This geographic distribution highlights the company’s bet on testing models in diverse work environments to feed motion algorithms before exporting them to global markets.

What actually changes for organisations and technical talent in the Gulf and Jordan?

This development signals a qualitative shift in the skill set required in regional labour markets; competition will no longer be limited to writing textual commands or traditional software engineering, but will open a direct path to new field professions linked to training physical systems, managing motion-capture crews and analysing spatial interaction. For startups, banks and regional industrial entities, the availability of local facilities for generating motion data reduces reliance on foreign infrastructure, lowers the cost of training automation solutions for operational environments in the Gulf, Egypt and the Levant, and enables testing robots in a setting close to real-world conditions without leaking sensitive data to non-sovereign platforms.