Portable Memory in IoT Applications
Feb 04, 2016
If I were to follow the boilerplate for an article that is talking about the Internet of Things (IoT), I would open with a prognostication from Gartner or Cisco stating that there will be X billion “connected” devices by 2020 or 2025. I’ll spare you. The fact is, however, that many companies are working on creating products that will be connected to the cloud, or they are working to connect existing products, thereby improving access to data so organizations can make better business decisions.
What role does portable or removable memory play in IoT devices? A recent article from the SD Association looked at that question through the lens of SD cards and stated applications of:
Supplementing internal memory by adding external removable memorySecuring data by using standard or proprietary security features of the removable memory
One of the more novel applications I have seen for using portable memory in an IoT application was from a Datakey customer. As with many Datakey OEM customers, this company used a non-volatile memory token for multiple applications.
Storing Location Information
The primary application was to store location ID information. With nearly all IoT applications, it is critical to associate the data being sent to the cloud with the device that produced it. For example, if revenue information is being sent to the cloud, it is important to identify and track which locations are generating the most revenue and which are underperforming. So why not just store this ID or location information permanently on the IoT device? Two reasons come to mind:
If the device has a minimal or no operator interface for data entry, the ID or location information will have to be entered when the device is provisioned (versus at the site). It is then critical for that device to be installed in its specified location otherwise the reported data will be tied to the wrong location. Alternatively, if a portable memory device is used, it can store the location/ID information and can be inserted on site. This especially comes into play with Reason 2.
Devices fail. When an IoT device fails, a new unit will need to replace it. If a portable memory device stores the location/ID information, transferring that information to the replacement device is as simple as removing it from the failed controller and inserting it into the new one.
Backing Up/Restoring Settings
When a device fails, having the location/ID information available on a portable memory device immediately allows the new unit to know who it is. Each device is likely programmed with a host of settings that are unique to its location. Once the device knows who it is (or where it is), it can then download this information from the cloud. The problem with downloading this setup information, often over a cellular network connection, is that the throughput is often not very fast and the customer has to pay for that data. If all of the settings are backed up to the portable memory device, the new device can quickly read them. The end result is the new device is a clone of the one that failed and it was restored quickly without pulling expensive data down from the cloud.
With the explosion of the Internet of Things, two of the major concerns that I see mentioned most often are a lack of standards and security. Portable memory can also be used for ensuring the connected device is authentic. Similar to how a SIM card is used to authenticate a cell phone to a network, a secure memory device (like a Datakey CryptoMemory token) can be used to authenticate an IoT device.
Specific applications may be able to leverage the portable memory device/port for other applications as well. If a device won’t connect and you need to reset it to its factory state or you need to manually apply an update to the firmware, a portable memory device can be an ideal way to accomplish these goals, especially if there are security or data usage concerns with doing these functions over the network.
Senior Product Manager - Datakey, ATEK Access Technologies, LLC