Most traditional models manage IT operations through manpower and static tools, accounting for the largest expense – approximately 40% – in managed services. Effective change, configuration and provisioning are all done through human intervention, which can be detrimental to efficient time and resource management.
However, "task based" or "workflow based" automation is being widely adapted for most routine tasks that are deterministic in nature, driven by standardised techniques or standardised operations. This has allowed IT teams to reduce human intervention to a minimum. A lot of IT employees' time is spent completing everyday tasks; depending on the individual's level of efficiency, this has a direct impact on performance and IT costs.
It must be remembered that these everyday tasks add no value to the business. What they do is maintain the current status and operation rather than developing IT solutions that service and move the business forward. Added to this inefficiency of operational maintenance is the possibility of human error during routine tasks. This can lead to further mistakes and more work, thereby taking more time away from the IT team that could be spent on tasks that add value.
As long as organisations are tied up with linear models they will continue to face operational challenges and administrative overheads. This is primarily because the complexity of the IT environment is increasing every day, and human efficiency is struggling to keep up.
Heuristic learning and IT automation
Whilst many businesses are looking to automate routine tasks that rarely change, this still involves a degree of human intervention, when automation needs to be reconfigured. A new approach to automation that incorporates "heuristic learning" can, and is, changing this.
Heuristic learning is a branch of artificial intelligence that uses mathematical algorithms and automation in order to "learn" how a process works. The idea behind it is to build analytical models that learn from data in an iterative fashion.
While current IT automation solutions are best suited to repetitive processes that do not change, as the corporate IT function is becoming so complex, automation based on set rules is no longer suitable or sufficient. Today, and in the future, approaches based on expert systems underpinned with heuristic learning are required as they can make decisions similar to a human administrator.
These new systems will be able to understand and process even complex decision-making processes. They can automate incidents, changes, releases or problems, with the ability to handle more complex issues such as capacity management or service continuity management without human intervention. With the ability to learn from past situations, intelligent automation systems will continue to develop and grow, benefitting the overall IT process and unburdening IT staff in the long-term.
Why businesses should look to adopt heuristic automation
Heuristic automation is a new concept that is not yet widely available. Nonetheless, businesses should look to invest in and take advantage of this technology where possible. Because its strength lies in knowledge-based automation, making it akin to a human administrator, over time all standard errors can be prevented proactively.
The modular and re-useable knowledge modules that make up a heuristic IT automation system can form work processes during an incident, and at the same time also create future processes for an unforeseen or unknown future error that is similar in nature.
Current IT automation standards are still high and help businesses to increase IT efficiency, lower costs and free up IT professionals' time to concentrate on more innovative projects that can add value to the business. As self-learning, intelligent, IT automation becomes more widespread, however, these benefits are only set to increase as the day-to-day running of the IT system will become a purely automatic, machine-driven process.
- Ramkumar Balasubramanian is General Manager – Automation and Practice at Wipro Ltd
from TechRadar
