Let’s Talk Tech #12: AI & Flooding, Developers & GenAI, and Spatial Computing

Let’s Talk Tech, the Orange newsletter for in-depth reporting on scientific innovations that will shape tomorrow’s world.

#PREDICTIVE #AI — Machine learning for flood rescue operations

Developed by Pranavesh Panakkal and Jamie Ellen Padgett at Rice University, OpenSafe Fusion is an automated data fusion framework to provide real-time information on levels of flooding on roadways, which is of vital importance to emergency services attempting to cope with natural disasters. The AI tool analyses and combines nine public data sources (video surveillance, social networks, sensors, etc.) to generate reliable predictions that can be further checked by humans. It is also collaborative, allowing fire services, police, hospitals, and other stakeholders to input data and better prioritize callouts.

#CODING — Keeping developers motivated in the age of GenAI

While generative AI has enabled developers to work faster, it can also engender new challenges by undermining effective collaboration and producing code that is overly homogenised. In response to these risks, CTOs and CIOs need to promote creativity and technical excellence and strike a balance when attributing tasks to humans and GenAI. Accountability, motivational workshops, and a long-term technological vision are essential to keep developers committed and avoid devaluing their skills. “Like any craftsmen, developers need to feel proud of what they produce,” explains Thibaud Courtoison, CTO of the start-up Prolong. “Every developer has different motivational levers,” adds Laurent Michelet, CTO of Shopopop, who organises special workshops to highlight individual motivational profiles.

#XR — Industry, real estate: the game-changing potential of spatial computing.

Now adopted by leading tech companies like Apple with its Vision Pro headset, spatial computing is transforming sectors such as industry, real estate, and entertainment with a range of immersive experiences such as collaboration using 3D holograms and virtual tours. With a combination of technological innovation and intuitive interaction, spatial computing is set to revolutionize the way we work, shop and play, paving the way for a hybridisation of the physical and digital worlds. Find out how it works and what it's used for with Morgan Bouchet, marketing manager for immersive and GenAI at Orange Group.

#NETWORKS — Towards the massive deployment of Open RAN

The new model for mobile access networks, Open RAN will pave the way for ground-breaking development in network architectures by facilitating the separation of software and hardware components and enabling greater flexibility and competitiveness. Favouring multi-vendor approaches, it reduces reliance on single equipment manufacturers. Orange has conducted promising experiments, including a pilot in Romania, which have demonstrated levels of performance equivalent to conventional RAN while reducing server power consumption by leveraging AI and the cloud. With the drive to capitalise on technical advances and reinvent operational models well underway, Open RAN will soon be ready for large-scale deployment.

#FOOD — New algorithm detects ultra-processed foods

A group of researchers, including Albert-László Barabási, a physicist renowned for his work in network medicine, have trained an algorithm, FoodProX, to predict the extent to which food products are processed. With its ability to detect ultra-processed foods, FoodProX can help consumers make healthier choices and contribute to better informed public policy. Now accessible online via the dedicated TrueFood website, FoodProX will open up new opportunities for targeted awareness-raising and food regulation initiatives.