The Physical AI Stack Is Being Assembled

gpcarracedoUncategorized

AI

The Physical AI Stack Is Being Assembled

By Gabriel Pastrana·July 4, 2026·4 min read·Issue #48

For years, companies competed by deploying robots. Today, the real race is to own the complete Physical AI stack, from perception and software to energy, data, and execution. This week’s announcements suggest the industry’s center of gravity has shifted.

If there was a common thread across this week’s news, it wasn’t another funding round or another robot launch. It was the realization that competitive advantage is moving away from individual machines and toward integrated systems capable of learning, adapting, and scaling.

Consider the sequence of announcements.

Symbotic’s acquisition of ARMS Innovations expands its ability to orchestrate warehouse operations beyond autonomous mobile robots. Packsize’s acquisition of Panotec strengthens its position in right-sized packaging by combining automation with packaging intelligence. Honeywell Technologies launched as an independent automation company, signaling a renewed focus on industrial software and automation infrastructure rather than diversified conglomerate priorities.

These are not isolated transactions.

They represent companies assembling platforms.

The same pattern appears in robotics.

Apptronik introduced Apollo 2 alongside a dedicated data collection and training facility. Figure continues expanding industrial deployments with BMW after successful pilot programs. AgiBot announced production of its 15,000th robot, highlighting how embodied AI is beginning to scale beyond prototypes. Rather than competing on hardware specifications alone, developers are investing in the data pipelines that continuously improve robot performance.

That distinction matters.

Large language models transformed software because they learned from enormous datasets. Physical AI will follow a similar trajectory, except its training data comes from warehouses, factories, construction sites, and distribution centers. Every pick, every inspection, every pallet movement, and every successful exception becomes another lesson for the system.

Perception is following the same path.

Luxonis closed a Series A round to accelerate its machine perception platform, reinforcing the idea that vision systems are becoming foundational infrastructure. Morph is embedding intelligence directly into robotic hardware through soft robotic cells, while Ambi Robotics and Pickle Robot demonstrated that customers increasingly value complete workflows over standalone robots.

Meanwhile, warehouse software continues its quiet transformation.

This year’s software survey, the Connected Warehouse feature, and the discussion around WMS, WES, and WCS all point toward the same conclusion: software has become the control layer that coordinates increasingly heterogeneous automation fleets. As facilities adopt multiple robot vendors, AI-driven planning, and real-time optimization, interoperability becomes a strategic capability rather than an IT project.

Energy is becoming another layer of the stack.

Lithium-ion systems are evolving from forklift batteries into enterprise assets that influence uptime, maintenance strategies, fleet utilization, and sustainability goals. As automation expands, reliable power becomes as critical as robot performance.

Even outside logistics, recent research from McKinsey & Company on semiconductors, AI infrastructure, solar energy, and pharmaceuticals reinforces the same message: tomorrow’s industrial leaders will be defined by how effectively they integrate hardware, software, energy, and data across entire ecosystems—not by excellence in a single technology.

The implication for supply chain leaders is significant.

The next competitive advantage will not come from purchasing the newest robot.

It will come from designing an architecture where perception, orchestration, execution, and continuous learning reinforce one another.

The companies building those architectures today are positioning themselves to improve every quarter—not because they buy more automation, but because their systems become smarter with every operation.

That is what Physical AI increasingly looks like in practice.

💰 Capital Is Flowing Toward Deployable Physical AI

Funding announcements from Luxonis, X Square Robot, Queue, and Morph share a common characteristic: investors are backing technologies that solve real operational problems. Capital is increasingly flowing toward perception, autonomy, healthcare automation, and robotic intelligence because these technologies can now demonstrate measurable productivity gains instead of future potential.

Investors are rewarding deployability, not promises.

🏭 Warehouse Software Is Becoming the Industrial Operating System

The latest warehouse software survey and discussions around the Connected Warehouse reinforce a familiar lesson: automation only scales when software can coordinate it. WMS, WES, and WCS are evolving into orchestration platforms capable of managing people, robots, inventory, and exceptions in real time.

As robotics fleets become increasingly diverse, software architecture—not hardware—will determine operational flexibility.

The warehouse of the future will compete on orchestration before automation.

⚡ Infrastructure Is Becoming a Competitive Advantage

Research from McKinsey on semiconductors, AI data centers, and energy points toward a broader industrial reality. AI requires physical infrastructure: chips, electricity, connectivity, and resilient supply chains.

Inside warehouses, the same principle applies. Lithium-ion fleets, sensing systems, wireless networks, and edge computing are no longer supporting utilities—they are strategic assets.

Physical AI depends on infrastructure as much as algorithms.

🌎 Sustainability Is Becoming an Operational Constraint

New packaging reduction legislation is pushing brands to rethink packaging strategies before regulations force change. Automated right-sizing, intelligent packaging, and material optimization are becoming strategic capabilities rather than sustainability initiatives.

Companies investing early will improve both compliance and operating costs.

Sustainability is increasingly shaping warehouse design decisions.

  • The Automate 2026 show reinforced one clear message: the industry’s focus has shifted from showcasing individual robots to demonstrating complete AI-enabled workflows.
  • McKinsey & Company argues that repeat builders consistently outperform first-time innovators because they institutionalize learning instead of relying on breakthrough ideas.
  • Built Robotics secured a $75 million contract, another sign that Physical AI is rapidly expanding beyond warehouses into construction and critical infrastructure.
  • X Square Robot reached a $2.8 billion valuation after four consecutive funding rounds, reflecting sustained investor confidence in embodied AI.
  • Queue is applying autonomous systems to pharmacy operations, showing that highly regulated industries are becoming fertile ground for Physical AI.
  • Carolina Handling and TinMan Systems partnered to bring predictive monitoring to conveyor infrastructure, extending AI deeper into existing facilities.
  • C.H. Robinson expanded its secure transportation capabilities for high-value freight, reinforcing that resilience and security remain strategic differentiators.
  • Recent research from McKinsey & Company on semiconductors, pharmaceuticals, solar power, and AI data centers points toward the same conclusion: infrastructure—not algorithms—is becoming the limiting factor for the next wave of industrial AI.
  • Humble Robotics believes recent advances in foundation models have finally aligned technology with the long-standing vision for autonomous freight.
  • Advances in robotic dexterity continue despite fierce competition, as Clone Robotics announced new funding following the settlement of its trade secret dispute with Tesla.

Five years ago, competitive advantage came from deploying better automation.

Today, it comes from owning the architecture that allows Physical AI to improve itself.

The companies leading this transformation are no longer thinking in terms of robots, software, batteries, or perception as independent technologies. They are assembling integrated systems where hardware, software, energy, and data continuously reinforce one another.

That shift may become the defining competitive advantage of the next decade.

The question is no longer:

It is:

Smart Automation

A four-minute weekly newsletter on automation, AI, intralogistics, supply chain, and operational excellence.

Subscribe free