The issue

In 2019, a Symantec security report found that around three quarters of all hacks were attempted on point-of-access devices such as routers. A mostly unsurprising fact given the scope of the access a hacker could obtain if he or she was able to compromise it, says Daniel Albertini, CTO, Anyline.

With that mostly common sense observation in mind, why do so many businesses still insist – in most cases completely needlessly – on sharing data via networks rather than simply processing data on device?

Think of scanning parts in a modern factory. An employee wants to add recently received stock into an inventory so he scans a serial number on the part using Optical Character Recognition (OCR) software on their mobile device to do so.

To scan the number of the part, that software is now sending data over the network to initiate the action of recognising the numbers – completing the process of scanning and digitising it. With the data that is being sent over the network exposed to any rogue agents who may have compromised the router being used.

So what?

While that case study may serve as a good example of the process, it is fair to say that most might argue exposing stock data is not the end of the world. So what if we turn to the far more sensitive topic: mobile scanning of identity documents.

Imagine a border control agency that has equipped its officers with a mobile solution to quickly scan passports and IDs. While this initiative may increase efficiency and improve accuracy vs. entering numbers manually to check documents, it runs into the same issue highlighted before.

To read the documents this software would have to send the information over Wi-Fi or mobile networks, thus exposing sensitive customer data to any hacker. This issue is then compounded by the fact that in order for your solution to work it must first send raw data outside of your organisation to be processed.

By sending information outside of your organisation, you could be putting your sensitive data in the hands of a third-party which may only be able to offer increased efficiency at such low costs because corners are being cut to do so in a competitive market. With cheaper solutions only offering OCR capabilities when connected to the internet.

The solution

In the examples above, these issues are caused by utilising software that is not capable of processing data on-device without needing a network. For some non-sensitive tasks this might be an acceptable trade-off for helping to speed things up at a lower cost, but when security is utmost, another solution is needed.

Implementing on-device processing can help to remove that need to send data away. When handling secure data, organisations such as border control can remain confident that the data is being processed on their own devices, with the added benefit of not needing a data connection.

While the benefits of utilising solutions such as Optical Character Recognition (OCR) are abundantly clear, the way it is implemented can have a huge impact on both the success and security of its use. Data provides organisations with power and OCR presents an opportunity to unlock further sources of that power through the digitisation of previously untapped information.

Investing in efficiency

OCR technology offers businesses huge opportunities when it comes to reducing human error, increasing efficiency and even reducing costs, but like all solutions, to truly gain value, businesses must invest in safe, class-leading solutions that do not cut corners and offer benefits without undue risk.

The author is Daniel Albertini, CTO, Anyline.

Comment on this article below or via Twitter: @IoTNow_OR @jcIoTnow