The future of AI is physical: The next frontier for Malta’s business leaders

The Deloitte Physical AI Dossier describes 40+ real-world applications available now across six major industries.

Over the past decade and more, artificial intelligence (AI) has transformed how organisations process information, analyse data, and make decisions. Yet despite its remarkable progress, AI has remained largely confined to the digital realm, optimising spreadsheets, predicting customer behaviour, and automating back-office processes. The next wave of AI is fundamentally different. It is moving at scale into the physical world, where intelligent systems can perceive their surroundings, reason about what they encounter, and act in real time. This shift from intelligence to action represents one of the most significant operational transformations available to businesses today.

Physical AI is technology that understands, exists in, and interacts with the physical world. Yes, it includes those things from our childhood imagination like autonomous humanoid robots, but it also includes next-gen sensors for active system monitoring, and advanced predictive analytics that drive real-world actions. Many economies are moving at speed with advanced AI applications. Malta needs to ensure it is keeping up.

Marc Alden, CEO, Deloitte Malta

Physical AI Is already here

Physical AI is not science fiction. It is already being deployed in factories, warehouses, hospitals, farms, and city streets across the globe. Autonomous mobile robots moving materials through warehouses with minimal human intervention. Drones inspecting infrastructure in places too dangerous or remote for workers to reach safely. Autonomous vehicles moving people across cities. Humanoid robots are being trialled for handling routine tasks in hospitals and care homes. Physical AI systems are becoming an increasingly accessible and economically viable operational capability.

What makes Physical AI distinct from traditional automation is its ability to operate in unstructured, dynamic environments. Traditional industrial robots require carefully controlled conditions and extensive programming. Physical AI systems can perceive and adapt to changing circumstances, navigate crowded warehouses alongside human workers, inspect complex infrastructure and identify problems without being explicitly programming, and learn from experience to improve performance. This flexibility opens possibilities unavailable to previous generations of automation technology.

Humanoid robots managing automated warehouse storage

Reimagining how humans and machines work together

For organisations, the promise of Physical AI extends far beyond incremental efficiency gains. It requires leaders to reimagine how work gets done and to intentionally design the relationship between systems and people. Rather than automating isolated tasks, Physical AI enables organisations to redesign entire operational systems, like how materials move through facilities, how maintenance is performed, how products are manufactured, or how services are delivered. When approached strategically, Physical AI can transform business models, optimising costs at scale, accelerating cycle times, and elevates people to higher value tasks. The organisations that treat Physical AI as a transformation opportunity rather than a technology procurement will capture the greatest value.

Implementing Physical AI at scale requires time, intentional design, purposeful retraining, and patience. Some tasks will be fully automated while others will still require human oversight and judgment. Most will involve collaboration between human workers and intelligent systems, with each contributing their unique strengths. This demands clear thinking about which tasks suit automation, which require human decision-making, and how accountability is defined when autonomous systems act on an organisation’s behalf.

De-risking deployment and governance

The foundation for successful Physical AI deployment is simulation and digital twins – virtual replicas of physical systems. Rather than deploying systems in the real world and learning through trial and error, organisations can design, test, and validate physical systems virtually before committing capital. This approach dramatically reduces risk, accelerates development timelines, and allows organisations to explore multiple scenarios. For Malta, where space is at a premium and operational disruption carries high costs, this simulation-first approach is particularly valuable.

Getting your governance framework right is critically important when deploying physical AI, as robots cannot be held accountable. Since these systems operate in the physical world, the consequences of failure are tangible and immediate. A robotic failure might injure a worker or damage equipment and to manage this risk we should give rigorous attention to robustness, safety, security, and reliability. Governance systems must ensure control and oversight is maintained, with clear human accountability structures so responsibility can be traced when something goes wrong. Transparency is essential so operators can understand why systems take actions and can intervene if necessary. These governance considerations are prerequisites for building the trust and confidence that allows organisations to scale Physical AI responsibly.

Opportunities across Malta’s key industries

Deloitte Malta’s industry leaders were asked to imagine how physical AI might be implemented in local industries

Malta’s business environment presents opportunities for Physical AI adoption. The country’s private and public sector organisations could all stand to benefit from intelligent physical systems. For Maltese organisations to adopt these technologies at the pace and scale needed to deliver tangible benefits, business and industry leaders must intentionally plan while actively learning from economies already operating these systems. Malta’s long-term competitiveness relies on steady technological advancement, appropriate risk taking and intelligent planning.

Robotic warehouse logistics and support

Consumer – John Debattista

“Imagine a warehouse manager watching the floor. Robots move materials autonomously, navigating around workers and obstacles, adjusting their routes in real time. Staff are freed from repetitive material handling to focus on tasks that require judgment and skill. The facility handles more volume without hiring more people.”

Autonomous inspection and defect detection

Energy, Resources & Industrials – Rachel Zarb Cousin

“Imagine a technician walking through an industrial facility. Instead of manually inspecting every component, robotic systems scan the equipment, identify wear and tear, and highlight areas that need attention. Extend this idea to our utility networks: an autonomous drone inspects energy infrastructure and provides live feedback on potential faults. Technicians can focus on the problems that matter. By targeting faults before they fail, we can ensure downtime is reduced in the factory or across the grid.”

Autonomous cash forecasting and ATM logistics

Financial Services – Ian Coppini

“Imagine an AI analytics system using real-time demand signals mixed with local real-world data such as weather forecasts, foot traffic, or even feast schedules to anticipate which ATMs are at risk of running low on cash. The system could automatically coordinate routes and prioritise drops for cash handling teams, directing them to the machines that need replenishment most urgently. Cash is always available when customers need it, and logistics costs drop because routes are optimised in real-time.”

Smart city traffic management

Government & Public Services – Marc Alden

“Imagine we deploy physical AI sensors across the island which detect and analyse congestion patterns and predict rush-hour bottlenecks. The system automatically adjusts traffic signal timing (where those exist) and sends routing guidance to drivers via an app, dynamically redirecting traffic away from congested areas. The result is smoother traffic, fewer bottlenecks, and a better experience for residents and visitors.”

Robotic support in eldercare

Life Sciences & Healthcare – Antoine Carabott

“Imagine a humanoid robot sitting with an elderly patient with dementia, engaging in conversation and providing companionship. The robot reminds the patient to take their medications at the right time and monitors whether they do. When doses are missed or concerning patterns emerge, the robot creates an alert. The robot provides continuous patient presence without fatigue, supplementing family and human caregiver presence. Patients feel less isolated, medication compliance improves, and caregivers experience reduced burden.”

Autonomous network monitoring

Technology, Media & Telecommunications – Craig Schembri

"Imagine a network technician receives an alert about a potential fault. Instead of spending hours guessing what’s wrong, the system has already combined network data with drone inspection footage to pinpoint exactly what is failing and where. The system automatically reroutes traffic away from the affected segment and contains the problem. The technician is then able to focus on resolving the issue. Outages will be shorter; customers experience fewer disruptions and technicians’ time can be better utilised."

What to do next?

These use cases are not distant possibilities. They represent applications that are being deployed successfully in other markets today and are ready for implementation. This dossier helps Maltese business leaders developing a real-world understanding of how physical AI is likely to shape their industries in the immediate future and help them actively plan for implementation of Physical AI in their businesses. Organisations that begin to explore these opportunities now, that invest in understanding the technology, assessing its applicability to their operations, and build both the data foundations and the governance frameworks necessary for responsible deployment, will be the ones that capture the greatest value. For support on understanding the applicability of Physical AI in your business, reach out to Deloitte Malta.

Yes

No

The future of AI is physical: The next frontier for Malta’s business leaders