Cognitive applications are some of the most powerful tools in today’s digital business landscape.

They transform mission-critical business systems by automating transactions and processes. Functions like campaign to lead, order to cash, procure to pay, incident to resolution, concept to market and hire to retire can now all be optimized and accelerated by artificial intelligence.

Autonomous enterprises that rely on cognitive applications are continually striving to automate these precision decisions at scale. They are driven by the need to reduce margins, technical debt and their investment in core systems. While traditional, transactional applications seem to have run their course, automation delivers game-changing benefits, including reduced staffing, fewer errors, smarter decisions, security at scale and better business outcomes.

How will autonomous enterprises evolve in the years ahead? The answer to that question begins by considering which decisions in the future will require human judgment versus intelligent automation.

The five levels of autonomous enterprises

According to Constellation Research, there are five levels of autonomous enterprises. Constellation predicts that by 2030, cognitive applications will be able to deliver full operational autonomy in this USD 10.35 billion market. Here is a quick overview of the expected timeline and functionality for each of these five levels:

Level 1: Basic Automation

Systems provide basic task and workflow automation.

• When: Today
• Includes: Basic process automation tools like BPM, manual instrumentation and control and intelligent workflow automation
• Who’s in control: Humans who guide many manual steps

Level 2: Human Directed

Enables human-directed automation of business processes.

• When: Current state of the art
• Includes: Robotic process automation, process-mining tools, journey-orchestration tools, machine-learning algorithms, natural language processing
• Who’s in control: Humans direct major decisions, minor decisions are automated over time with some effort in training

Level 3: Machine Intervention

Delivers automation with occasional machine intervention.

• When: The next big thing in 2021
• Includes: Cognitive applications, neural networks, GANS models, contextual decisions and next-best actions
• Who’s in control: Humans still on standby, but can be hands-off for periods of time

Level 4: Fully Autonomous

Presumes that machines can deliver full automation but not sentience.

• When: 2023
• Includes: AI-driven smart services, full automation, self-learning, self-healing and self-securing
• Who’s in control: Machines fully automated

Level 5: Humans Optional

Full sentience and humans may no longer be needed.

• When: 2030
• Includes: Fully autonomous sentience, empowering precision decisions at scale
• Who’s in control: Humans fully optional

The bottom line: Level 4 autonomous enterprises to emerge by 2023

According to Constellation, pioneering work with early cognitive applications shows exponential progress and may achieve Level 4 status by as early as 2023. This means organizations will have to rethink how they work with their transactional applications. There will also be a major impact in how businesses should approach future data-driven digital networks and distributed compute and storage environments.

For most enterprises, it’s clear that the future will be autonomous. Machines will deliver services that are continuous, auto-compliant, self-healing, self-learning and self-aware. An organization’s need for greater precision decisions will require new connections to data-driven digital networks and more and more sources of data. Ultimately, the battle for public, private and shared data will shape who wins in the new networked economies that will form the future of this autonomous decade.

Ready to build your autonomous enterprise? Discover how you can automate business and IT processes at scale with IBM® Robotic Process Automation.