We’ve all heard the hype about how autonomics — automated, self-managing processes — will revolutionize IT infrastructure services and, by extension, the supply chain and manufacturing, warehousing and distribution. All too often, when we hear these kinds of promises, we’re disappointed when new automation solutions are introduced with much fanfare but with no concrete customer references to demonstrate real-world effectiveness.
However, this time might just be different. Experts agree that autonomics has the potential to unleash a significant wave of service improvement and cost-efficiency across a variety of industries. One thing is for sure: IT infrastructure autonomics is inevitable, and those of us in the trenches need to prepare for it. The sooner the better; autonomics will reduce service faults and labor costs. In fact, we are likely to see as much as a 40% to 75% reduction in labor cost from automation technologies that include autonomics, versus the 15% to 30% savings we have gained over the years through labor arbitrage.
At the extreme, autonomics includes cognitive computing in which software can process natural language commands and respond based on defined contexts or goals. However, those in the infrastructure world are most likely to first exploit the self-healing capabilities that autonomics enables: some level of automatic fault discovery and ability to correct those faults. As a result, IT infrastructure will be able to automatically adapt to changing conditions and produce service quality and reliability above and beyond current infrastructure automation.
While cognitive engines are rapidly coming to market and are successfully being used in “ideal” environments, such as cloud providers’ data centers, they aren’t yet ready for “prime time” at the enterprise level. Commercial, off-the-shelf autonomic products that address the lowest-hanging infrastructure fruit (think server management) are already on the market and are very much a part of solutions being currently proposed by service providers.
In the next 12 months, we can expect to see impressive enterprise-level success stories that demonstrate how autonomics, at a minimum, can greatly reduce automation failures and the need to update automation playbooks.
Here are the Top 5 steps IT leaders in manufacturing and industrial enterprises can take now to get ready:
1. Educate yourself on autonomic technologies. It’s important to distinguish between traditional IT automation (a set of scripts to handle anticipated events) and autonomics (software that can learn). IT automation has been in the marketplace for years and has proved to be cost effective when processes are highly standardized. Autonomics software is just now being introduced in “friendly” client environments in which the relationship with the service provider is healthy enough to withstand experimenting with new services.
2. Don’t try to do it yourself. It might be tempting to embark on a plan to internally develop autonomics for your infrastructure, but both the costs and risks are high. Service providers gain from economies of scale and experience that a buy-side organization cannot approach; they focus on use cases for highly repeatable infrastructure processes that can be reused across customers. While traditional automation, like scripting, is worth pursuing internally or via service providers, do-it-yourself autonomics would be like writing your own server operating system.
3. Standardize and homogenize your infrastructure. Simplification has been and continues to be the Holy Grail. Move your businesses off their legacy platforms. Even if your organization can only focus on a subset of your overall environment (one geography or one business unit), keep on your path or accelerate it. This will provide a platform when autonomics is ready for prime time and, in the short term, it will lower support costs and increase reliability. As we’ve learned over the last several years, standardization increases agility and accelerates the adoption of other emerging technology.
4. Continue IT service management process deployment and maturation. Robust, well-documented processes are the starting point for automation and subsequent autonomics. Focus on development of “light” processes that enable consistency and agility. Autonomics will help your organization deal with regional and business unit variations — though more process homogeneity is better than less.
5. Develop and implement an intellectual property (IP) strategy. A key automation risk is getting locked into one provider if it owns the automation IP, the automation systems and/or the instructions put into those systems. Your IP strategy and legal approach has to account for outsourcing arrangements as well as best practices regarding ownership of software licenses and other IP.
Steve Follin is a Director in ISG’s IT sourcing practice, bringing more than 25 years of experience in information technology sourcing, IT solution design, and IT strategy. He has proven expertise in IT infrastructure sourcing, cloud computing, end user computing, data center consolidation, IT assessments, and optimizing IT costs and efficiencies. He has developed and led programs for technology sourcing, technology infrastructure and service quality improvement, cost and operational improvement, IT governance, and IT program management. His functional expertise includes IT sourcing advisory services, infrastructure support and transformation, cloud computing, IT strategy, and enterprise architecture.