Operating in contested or denied environments isn’t an AI issue, it’s an integration and infrastructure challenge.
By: Dave Dimlich
President of SD3IT
It may seem like artificial intelligence is everywhere these days, but for military and commercial personnel working in remote environments, it can seem like it’s never there when you need it.
Military personnel, emergency responders, critical infrastructure operators and industrial organizations frequently operate in environments where the cloud, so ubiquitous in most of society, is a mere rumor with connectivity being limited, intermittent or unavailable altogether. At the tactical edge, waiting for data to travel to the cloud and back isn’t just inefficient. In some settings, it’s impossible. And even sporadic connectivity can be the difference between mission success and failure.
Efforts to make use of AI in disconnected, intermittent or limited (DIL) environments created by distance, terrain or electronic warfare compounds the challenge. AI tends to require copious amounts of processing power, which in most cases occurs in data centers connected via the cloud. But in deployed environments, organizations in a variety of sectors are looking for ways to localize AI processing, having it perform analysis or other steps on the spot.
The challenge isn’t AI. It’s performing the integration and developing the architecture necessary to allow AI-powered edge devices to work together, allowing personnel to make decisions and act even when isolated from centralized control.
The Benefits of Processing Data At The Source
The tactical edge generally refers to forward-deployed environments where people and systems must make decisions in real time despite limited communications. Although often associated with the military, the same challenges exist outside defense. A utility company restoring power after a hurricane can’t assume constant connectivity. A physician using diagnostic equipment in a remote location may not have access to cloud resources. Manufacturers monitoring automated production lines and performing predictive maintenance can’t afford delays while critical sensor data travels across the country for analysis.
Organizations are increasingly looking to solve that problem by making decisions where the data is generated, which is where edge computing and localized AI become essential. Instead of transmitting enormous amounts of sensor data back to centralized servers, AI processing occurs directly on ruggedized edge hardware, whether that’s a vehicle, drone, wearable device or localized compute platform. Only essential information needs to move across the network.
The benefits to deployed units are significant. Local processing reduces latency, allowing systems to react in real time. It improves resilience when communications are degraded or unavailable. It also reduces bandwidth requirements while potentially strengthening data security because sensitive information doesn’t always need to leave the local environment.
For military operations, that means commanders maintain situational awareness even when communications are disrupted. For industrial organizations, it means predictive maintenance systems continue monitoring equipment despite network interruptions. And for healthcare providers, it means diagnostic tools can continue assisting clinicians in remote environments.
The Challenges of AI in Remote Settings
But that approach can be difficult to pull off. Edge devices often must process complex processing in real time while staying within strict power limitations. They can’t rely on the kind of fast data retrieval they get in cloud-supported settings and must operate in unpredictable environments where physical damage is always a possibility. They also must operate within a fragmented edge ecosystem where devices in use aren’t always compatible.
If that’s not enough, operating at the edge also has become a focal point in the cybersecurity wars, in part because edge devices are often less protected than other network endpoints.
A number of defense and commercial projects are trying to tackle the issue. The Army in March launched Project ARIA, or Army Rapid Implementation of Artificial Intelligence, to develop tactical-edge AI solutions and get them into the field in months rather than years. The service also has a number of other initiatives underway under the umbrella of Project Linchpin, its flagship program to develop a distributed edge AI architecture.
Other examples include the Marines’ Project Dynamis, which aims to develop next-gen, AI-powered command and control capabilities for contested environments with a mesh network connecting sensors, systems and data across multiple domains. The Navy, meanwhile, is working to deploy AI strike teams that use AI and machine learning at the edge while adhering to zero-trust data security principles.
Building an Ecosystem for the Edge
These latest initiatives, along with private-sector efforts such as those focusing on infrastructure or autonomous robots, recognize that deploying operational AI involves a lot more than installing an AI model on a device and turning it on. It requires an entire ecosystem working together.
Edge devices have limited computing power, along with strict size, weight and power requirements. And they must often be able to withstand harsh environmental conditions. Devices need to be optimized to run efficiently without sacrificing performance while working in environments with unreliable communications and maintaining protection from cyber threats and physical tampering.
The key challenge, however, comes with getting all those devices to work together. How do you update thousands of distributed AI systems? How do you monitor performance across widely dispersed locations? How do you ensure every deployed model remains secure, trusted and up to date with current data?
Those questions have less to do with artificial intelligence than with infrastructure, integration and lifecycle management. Successful edge AI depends on resilient networking, rugged computing platforms, secure identity management, zero-trust security, data orchestration and centralized visibility across distributed environments. Without those foundational capabilities, even the most advanced AI models struggle to deliver reliable operational value.
AI at the edge requires that organizations understand where systems will operate, what constraints they’ll face and how they must perform when conditions are less than ideal, which is the focus of SD3IT and its partners. As a value added reseller and integrator for federal, military, infrastructure and other companies, we work with partners such as edgeTI, for example, on ensuring that disparate edge systems can operate under centralized control with platforms such as edgeCore.
The goal is to help customers integrate the technologies that make operational AI possible, bringing together rugged edge computing platforms, resilient architectures, secure networking, zero trust security and centralized management into a unified environment that allows organizations to monitor, orchestrate and manage distributed resources from a single operational picture.
AI’s Future at the Edge
AI will continue to reshape defense, government and commercial operations for years to come. Bringing operational AI to the tactical edge is one of the most significant shifts now underway, particularly for organizations operating in remote, rugged or communications-constrained environments. In those settings, localized processing, resilient architectures and real-time decision-making become strategic advantages rather than technical conveniences.
AI’s future will increasingly be found at the tactical edge. But operational success won’t depend on the sophistication of the AI model. It will depend on building resilient architectures, trusted integrations and secure infrastructures that allow AI to operate wherever the mission takes it.
About SD3IT
SD3IT (Solution Driven, Designed and Delivered Technology) helps federal agencies, defense organizations and commercial enterprises integrate secure, resilient technology solutions that support mission success. By combining deep technical expertise with trusted partner ecosystems, SD3IT delivers infrastructure, cybersecurity, cloud, edge computing and AI solutions designed for real-world operational environments.
