Edge Computing: From Big Data to Smart Data

As technology advances, businesses are amassing more and more data. With the advent of cloud computing (to Edge Computing), and with more servers and greater bandwidth than ever before, we now have the ability to both generate and store enormous volumes of it. However, it’s one thing collecting all this data and another knowing what will actually provide useful, timely insights for us.

The goal is always to derive insight from data in a cost- effective way and, with the emergence of edge computing, businesses are increasingly able to process data where it is generated – to decide what is valuable and what is not at source without having to convey it, at considerable expense, to a data-processing warehouse first. This is both cheaper and more efficient than transferring a constant flow of data, much of which may be completely irrelevant to a company’s needs.

Edge computing and cheap sensors are fueling the rise of predictive maintenance.

_{Frank Sauber is director of ecosystem partners at Sigfox}

In the context of the Internet of Things (IoT), sources of data generation are usually “things with sensors” or embedded devices. For example, a smoke detector might be constantly transmitting data between itself and a central system, but the ability to access and analyze that data only matters if something is wrong – if the device starts detecting smoke. At that point, you want a lot of data or information so you can assess the situation quickly and take the appropriate action to get things back on track.

The goal is always to derive insight from data in a cost- effective way and, with the emergence of edge computing, businesses are increasingly able to process data where it is generated – to decide what is valuable and what is not at source without having to convey it, at considerable expense, to a data-processing warehouse first. This is both cheaper and more efficient than transferring a constant flow of data, much of which may be completely irrelevant to a company By installing a simple, secondary system that remains dormant and only transmits data or wakes up the rest of the system when the need arises, businesses can be far more efficient and cost-effective in their practices.

Predictive maintenance

Edge computing, together with the decreasing cost of sensors, is leading to the rise of predictive maintenance – the ability to determine the condition of equipment and estimate when maintenance should be carried out. With tasks performed only when needed, rather than imposing regular “just in case” checkups, businesses can save time and money by not taking perfectly good equipment out of circulation.

A low-cost sensor could be connected to a low-power network and placed on a motor, for example, with an alarm programmed to trigger if a certain temperature is exceeded, or when the motor has reached predefined operating hours. Shock detection could also be factored in to calculate impacts on defective bearings. Once the alarm is triggered, either a technician could be sent to inspect the motor, or the system could activate the sensor over broadband communication and request a complete monitoring log to allow for more detailed analysis.

Giving Peace of mind

For businesses, operating a complementary communications system is not only more cost-effective in the long term but can act as a backup for primary systems in case of a breakdown or security breach. A backup system can sense when the primary system has been compromised – be that hacked, jammed, or out of power – and kicks in to keep the service up and running.

For example, Freebox, an Internet set-top box from French company Free, uses IoT service provider Sigfox to remotely diagnose millions of its boxes for basic failures, to ensure a more reliable service and a better customer experience. It also provides a backup channel if the Internet connection goes down and additionally helps customers during the installation phase, as well as being a source of backup information during IP backbone maintenance.

Network of Security: Professionally monitored home alarms like Verisure’s Securitas Direct connect services for smart homes using Sigfox’s global network for backup and antijamming. Because the system does not run over the GSM network, it can kick in at this point to ensure the service doesn’t go down.

Additionally, Verisure, a security company that provides professionally monitored home alarms and connected services for smart homes, uses Sigfox’s global network to back up the company’s home alarm systems and to provide anti-jamming capabilities. The alarms are connected to a central system using a GSM phone link but savvy thieves can sometimes use jammers to scramble the GSM signals and deactivate the alarms. Because Securitas Direct’s backup system uses Sigfox and not the GSM network, it can kick in at this point to ensure the service doesn’t go down. This not only brings peace of mind to customers at a low cost but provides a competitive advantage to Securitas Direct.

Enabling Better access

A secondary system that runs on a different network can also help to enable better access for different stakeholders. In a smart building there might be a local area network (LAN) but if it’s owned by the building operator, not the building owner, communication might be there but not accessible to everyone.

In the case of building maintenance communication systems might be accessible to building operators, but not to maintenance staff or cleaning personnel, through a digital subscriber line (DSL), with solid protection through firewalls preventing access to straightforward information like temperature or humidity. By placing low-cost sensors onto objects and creating a secondary network, maintenance staff can monitor environmental conditions without needing to be granted access to the primary system. This helps them to diagnose problems faster and keep everything running smoothly.

Employing simple, low-cost, and complementary communications technologies that only transmit data if and when needed means they can also act as a backup for existing systems or simply work alongside them. In the coming years, we are likely to see an increasing number of companies bringing intelligence closer to devices, filtering data to a far greater extent before sending it to the cloud or server farms, and an increasing demand for “smart” data where collection is “as needed”. The future lies in businesses having a mix of communication technologies, employed for different purposes, that can be accessed by different owners. This will not only provide them with greater flexibility and security but be much more cost-effective, too.

READ MORE ARTICLES