Open standards essential to Cloud

OvumWith a healthy dose of sharp-edged showmanship – describing Salesforce as the “roach motel” of clouds – Larry Ellison launched Oracle’s public cloud service.
While we believe that utilities may start adopting cloud services to help manage smart meter data, particularly for early-stage trials, it is unlikely that any will favor public clouds over the more secure private equivalent. However, utilities selecting cloud vendors should heed Ellison’s warnings of how proprietary standards in clouds complicate the movement of data and applications from one cloud to another, or to an enterprise’s own data center.
To date, very few utilities have adopted cloud services for data storage or hosting applications, for a variety of reasons. For example, on the operations side of the business, data created by control systems are incredibly valuable: there is significant risk to storing data offsite, and hosting mission-critical applications such as outage or distribution management systems in the cloud will be anathema to almost all utilities.
On the retail side, historically, there has been little reason to move customer data to the cloud: it remains fairly constant, and relatively low in volume.
However, with the deployment of smart metering projects across North America, Europe, and beyond, utilities face exponential growth in customer and operational data. While old electromechanical meters may be read between once a month and once a year, smart meters can be configured to read far more frequently – typically every 15 minutes or 30 minutes, although some utilities read at even shorter intervals. In addition, smart meters can monitor power quality and send out other alerts. This operational data gives distribution network operators the potential to gain far better insight into what is happening at the edge of the network.
But smart metering does not come cheap. Its costs include not only the metering hardware and communications technology to transmit meter readings, but also the huge infrastructure transformation that the utility must undertake to cope with storing meter data and billing based on granular consumption data. It should come as little surprise that the vast majority of first movers in smart metering were larger utilities with balance sheets that could bear the risk of investing in largely untested technologies. Without exception, smart meter data remained on these early movers’ premises.
Today, smaller utilities are deploying smart meters, particularly in the Nordic region and Germany. These utilities are also more likely to sign up to transaction-based metering services from third-party providers rather than take on the entire project risk. These providers now offer metering services to much larger utilities that are just embarking on early-stage smart meter trials. Rather than invest upfront in a smart metering platform for its early-stage trial, the utility can use a cloud-based service to trial different technologies. The utility can then continue to use the cloud service during, or after, full-scale deployment of smart meters, or move data and applications back on-premise at any time.
Ellison warns of false clouds
Using [an] acrimonious dispute with Salesforce as a backdrop, Oracle CEO and chairman Larry Ellison launched his company’s public cloud at Oracle OpenWorld 2011.
He stressed security, noting that Oracle’s cloud is based on logically partitioned virtual machines, as opposed to Salesforce’s multi-tenant architecture. In addition, Ellison was keen to promote how the security is embedded in the middleware of Oracle’s public cloud, rather than in applications themselves. There will, no doubt, be lively debate over the pros and cons of virtual machines versus multi-tenancy. Ellison’s keynote focused on Oracle’s public cloud offering. However, few utilities will be prepared to commit their customer data to a public cloud.
Oracle’s cloud computing strategy also includes an option for private cloud: all cloud-based services adopted by utilities, at present, are on service providers’ private clouds, and this trend is likely to continue.
Regardless of the model – whether a public or private cloud – the industry will do well to heed Ellison’s warnings about his rivals’ cloud services. Ellison warned about false clouds, whose lack of open standards makes it difficult to move data and applications from one cloud to another, or back to an enterprise’s own data center.
Utilities using cloud services for smart meter trials should keep their options open with regard to their long-term data management strategies. Once a deployment has completed, smart meters will create relatively constant and predictable volumes of data. While cloud-based data services will always remove the requirement for upfront capital expenditure in data center infrastructure, the benefit of variable capacity on demand does not apply to utilities. On-premise management of smart meter data may prove more economic than cloud-based alternatives.

Original article: Utilities evaluating cloud services should avoid the “roach motels”