Product & Tech
Seb
At MODEX 2026, Exol launches U.S. "physical AI" facilities and a fulfillment-as-a-service model — embodied agents and autonomous handling running live inside last-mile fulfillment nodes
Use: "physical AI facility" is now a productized category, not a research term — and every embodied agent inside it is starved without clean, reconciled field data. Position daXos as the ground-truth feed those facilities run on, and watch fulfillment-as-a-service as a channel where our layer ships embedded.
Physical AI in logistics and manufacturing forecast to reach 400,000 deployed systems by 2030 — sensor-fed AI whose decisions trigger physical action moves from pilots to fleet scale
Watch: 400k embodied systems is 400k consumers of trustworthy real-world data. The TAM for our layer scales with this number, not with device sales — keep the narrative pinned to "every physical-AI system needs a governed data source," and use the figure in the deck.
AT&T, with Cisco and NVIDIA, pushes a network-driven edge-AI approach that folds compute, connectivity and zero-trust security into a single unified architecture for IoT
Signal: the carriers are unifying compute + connectivity + security, but stop short of unifying the application data across vendors — that gap is ours. Frame daXos as the data-unification layer that sits on top of their unified pipe, not a competitor to it.
IoT Analytics calls 2026 the inflection point for edge-AI IoT devices: OEMs move from pilots to broad portfolio refreshes marketed as edge-AI-enabled
Signal: a wave of smart-but-siloed endpoints is about to flood the field as OEMs refresh portfolios. The more endpoints reason locally, the more value migrates to whoever normalizes and reconciles them — keep the roadmap on provenance and cross-vendor unification, not raw collection.