AI in action: Robotics interoperable operations at the edge

In years past, the mention of artificial intelligence (AI) might have conjured up images of sentient robots attempting to take over the world. From the ruthless VIKI in I, Robot to the powerful cybernetic antagonist from Age of Ultron, fictional automatons perpetuated the notion that AI may unleash disastrous consequences.

However, both robotics and AI aim to augment human capabilities and improve lives. For example, robotics have long played a significant role in the industrial sector at the edge, from discrete manufacturing to continuous batch processing and hybrid manufacturing. With recent advances in AI, industrial organizations can enhance agility, flexibility, and interoperability in robot-to-robot and robot-to-human interactions. We call this robotics interoperable operations (RIO).

Five critical success factors

If your organization is looking to meld AI with robotics, success depends on five critical factors:

1. Deploying IT infrastructure that supports AI at the edge
2. Verifying edge infrastructure is resilient, secure, observable, explainable, and manageable
3. Establishing a centralized operating platform for application orchestration and management
4. Making sure applications are interoperable and reliable for robot-to-robot and robot-to-human collaboration
5. Maintaining data protection and security of RIO infrastructure and robots

Three core deliberations

To address these success factors, organizations should engage in three core deliberations, known as the 5-6-7 deliberation principle. These include five modes of business benefits, six types of AI-based applications, and seven foundational infrastructure considerations.

Dell

Five modes of business benefits

Start by considering the value that these business benefits can impart:

1. Increased flexibility and robustness with “human in the loop” systems leveraging human input and oversight to optimize human/robot collaboration
2. Improved efficiency via real-time process control for workflow optimization with manageability and configurability of individual robots
3. Better reliability and maintainability of robots, improving overall production uptimes
4. More agility in dynamic job planning and scheduling with humans in the loop
5. Enhanced safety, quality, and sustainability metrics from context-aware RIO

Six types of AI-based applications

These application categories drive the five business benefits:

1. Machine learning (ML) enables robots to perform tasks autonomously using model-based inferencing based on real-time data.
2. Natural language processing (NLP) helps humans interact naturally with robots via dashboards, text, speech, or speech-to-text.
3. Machine vision gives robots the ability to use sensor and camera data to interpret context-aware information.
4. Generative AI (GenAI) leverages ML and NLP to create new content, enhancing human-to-machine and machine-to-machine interactions and robot task execution.
5. Agentic AI achieves predetermined goals by interoperating between groups of robots or between robots and edge-based applications of manufacturing operations management.
6. AI-enabled digital twins create simulations from real-time data to visualize and control robots for optimizing task execution.

Seven foundational infrastructure considerations

Finally, your team will need to consider the infrastructure requirements for deploying and managing your selected AI-based applications.

1. Performance: RIO requires low latency for real-time decisions and high throughput via hybrid edge / on-premises processing for large volumes of data. Systems need to provide reliable performance even in tough environments.
2. Scalability: Compute resources must adjust elastically based on workload demands. Increased workloads from robots may require horizontal scaling, while supporting complex AI applications may require vertical scaling.
3. Storage: Data-intensive AI workloads require techniques for handling large data sets, including compression and deduplication. Keeping data close to where it is generated reduces access times, while distributed storage enables quick access and redundancy.
4. Networking: Stable, high-speed connections between robots and edge devices are critical, and network topology must support efficient communication and data transfer, with bandwidth optimization to help prevent bottlenecks.
5. Security: Data needs to be protected in transit and at rest with strict access control policies that can prevent unauthorized access. AI and ML can identify and mitigate security threats in real time.
6. Infrastructure integration: Different systems and devices supporting RIO need to work together seamlessly, using APIs and middleware to enable communication between components. Leveraging solutions that follow industry standards facilitates integration.
7. Robot mobility: Robots performing tasks in various process areas need mobility support, and systems need to provide continuous connectivity as robots move between network zones. Adaptive algorithms can optimize performance based on the robot’s location and movement.

Getting on board with RIO and edge computing

RIO is a transformative initiative for the industrial sector. Recent advances in technology — from sensors to communication technologies to GenAI — enable using robot data for advanced analysis, optimization, and collaboration, providing more flexibility, agility, and resilient production operations. If your organization is considering using AI to link robots, cobots, humanoids, or uncrewed vehicles with humans in the loop, then you are already on your journey towards robust process control and dynamic process optimization.

Learn more about how Dell Technologies can support your edge and AI journey.