In February 2019, an article published in Homeland Security Today, “Communications Convergence: What Does ‘Communications Convergence’ Mean for Border Security? 3 Ways to Prepare for LTE, LMR,” described strategies to prepare your organization for the transition from LMR voice communications to an LTE communications infrastructure with more capacity at a reduced total cost of ownership.
The strategies and tactics described in that article remain true and part of the overall program management solution. What is different is the continued push of surveillance “devices” closer to the source of criminal activity and away from centralized processing putting pressure on communications networks. It is increasingly clear that program managers will have to manage at the intersection of edge computing, artificial intelligence and communications. They will have to manage the risks that are introduced while managing the workforce as artificial intelligence assumes more tasks. The purpose of this article is to describe this “intersection” and propose ways in which to manage it.
Trends in Industrial Edge Computing
Centralized processing can increase latency and may impact law enforcement response time. Often data transfer and latency is the result of minimal to no communications infrastructure in remote locations where criminal activity takes place. In other cases communication infrastructure has not kept pace with dramatically increased requirements to transport large amounts of data. Using a transportation example, we have a lot of traffic getting off the highway (Edge) into single-lane off-ramps (Data Backhaul). There may be a mismatch between the collection of massive amounts of data and networks’ ability to transport it. This is simply the side effect of programs acquiring their communications specific to their requirement and not an overall communications design and architecture. There is already an impact to law enforcement operations that is expected to worsen unless an enterprise communications strategy, design and architecture is developed and deployed. This matters as industry experts expect industrial edge computing demand to increase, putting pressure on networks and their transport layers.
A Gartner article published Oct. 3, 2018, states, “While edge computing isn’t new, it’s beginning to take hold in the industrial sector – and the opportunity is far greater than anything we’ve seen in the consumer sector, and here’s why: Real-time data in a real-time world: The edge is not merely a way to collect data for transmission to the cloud. We are now able to process, analyze and act upon the collected data at the edge within milliseconds. It is the gateway for optimizing industrial data. And when millions of dollars and human lives are on the line, edge computing is essential for optimizing industrial data at every aspect of an operation.” This same article states, “Around 10% of enterprise-generated data is created and processed outside a traditional centralized data center or cloud. By 2025, Gartner predicts this figure will reach 75%.”
The demand to deploy more sophisticated “smart surveillance devices” will increase but it will have to be managed carefully as more tasks are supported by artificial intelligence software. This will require clear definition of operational requirements and mission priorities. We expect AI to continue to press organizations to define the line between “human-supported” and “AI-supported” activities. How far into the mission-essential tasks can AI be deployed when people simply must perform the task but AI can assist or when is the opposite possible or appropriate?
Edge, Communications and Artificial Intelligence (AI) – managing the intersection
Generally, each of these trends (edge, communications and AI) is challenging to develop, deploy and then integrate into potentially conflicting operational requirements and priorities. The expectation for program managers is to deploy their programs, get technology fielded and leave items outside of the scope of work to someone else. With increased pressure on budgets, it is likely program managers will look to consolidate activities and share existing civil infrastructure.
The following three strategies can move organizations toward improved management of these trends.
1. Develop and Enterprise Communication Strategy
- Develop an inventory of communication infrastructure assets
- Evaluate capacity of this infrastructure against operational data requirement
- Understand the deployment plans of “edge” devices and how much data they consume and transport
2. Evaluate Program and Acquisition Management
- Special emphasis on the Integrate Master Schedule with emphasis on integrated on intersecting project points between AI, Edge and Enterprise Communication Activities
- Obtain agreements between project managers on when to conduct integration so not to disrupt schedules, contracts or budgets
- Conduct engineering analysis to understand implications to integrating systems
- Determine impact to operations and sustainment of systems
3. Determine degree of Artificial Intelligence
- What are the operational requirements?
- What are the organization priorities?
- How much of the critical tasks can or should be automated with AI software?
We are at an interesting time in the adoption of artificial intelligence, the use of edge devices, data analytics and the impact on communications infrastructure. There is significant program risk to cost, schedule and system performance, but managing the intersection of these trends is inevitable. The ability of programs to take advantage of AI and concurrently plan for high capacity networks to transport information to where law enforcement needs it will be leading the way.