Every day, IT teams face productivity loss due to hardware failures in the data center. If a part is unavailable for the repair, longer downtime ensues. Waiting adds frustration and uncertainty to your day. Our OnDeck Predictive Sparing system avoids these kinds of parts slowdowns with its data-driven parts management capabilities.
OnDeck Predictive Sparing works proactively to reduce downtime
By analyzing 20+ years of parts, equipment and service data, OnDeck can predict failures and keep parts ready on the shelf for quicker, more reliable support. OnDeck drives our parts management strategy by identifying part needs and automating our supply chain and inventory sparing system. It proactively customizes local inventories with the right parts before a hardware failure occurs in a customer’s data center. This dynamic system continuously learns and adjusts parts inventories based on a steady stream of data from over 200 daily service events.
How OnDeck Predictive Sparing combines data, AI and automation
1
20+ years of historical service ticket and customer equipment profile data fuels the OnDeck® database.
2
Our proprietary algorithm analyzes usage trends to predict potential failures before they occur.
3
OnDeck® identifies the parts needed to support your equipment before a system goes down.
4
The correct parts are shipped to your local Service Express office or stocking location.
5
When you need service, your local engineer has the right parts in-hand to repair your equipment.
6
Every 30 minutes the database updates and machine learning analyzes current data for part needs.
Predictive sparing increases parts knowledge
Many companies claim they can accurately stock high-fail parts, but their methods rely heavily on “gut instinct” rather than using advanced algorithmic systems with historical evidence. Truly managing part demands requires a more sophisticated solution. Our advantage is understanding how parts work together and knowing the vulnerabilities of the tens of thousands of equipment types we support. While a power supply may often fail in one model, it can prove exceptionally reliable in another, and the data gives us those insights.
Anticipating part issues using dimensionality
By storing multiple dimensions of parts in our database, such as OEM, summary description, detailed description, critical/non-critical, etc., OnDeck can identify part needs for equipment that has yet to experience a failure. The system makes recommendations based on population sizes and historical rates for similar items. For example:
Making connections
Our multivendor database stores multiple dimensions of metadata for thousands of SKUs in the install base, including similar and alternate parts used across models and OEMs.
With this information, OnDeck can determine (without any historical evidence) that a fan in a new machine will likely fail if its design is similar to that of a previous generation system (based on historical evidence), and stock the part accordingly.
Ready for the next failure with OnDeck Predictive Sparing
Our expertise, extensive service database and OnDeck capabilities help us understand better than anyone all the interrelated dependencies of enterprise IT infrastructure. We know who produces the equipment, how it’s put together and how components, like memory modules, CPUs, hard drives, tape drives, etc., behave. Most importantly, we understand how parts fail, anticipate when they’ll fail and prepare for the failure.
The intelligence, data and AI housed in the OnDeck Predictive Sparing system ensures parts availability to help us meet customer needs at critical moments. Our first-trip repair and SLA-met rates are 97% and 99%, respectively.
Topics:
