By {{Article.AuthorName}} | {{Article.PublicationDate.slice(6, -2) | date:'EEEE, MMMM d, y'}}
{{TotalFavorites}} Favorite{{TotalFavorites>1? 's' : ''}}
Some of the most sensitive content a construction company deals with is related to jobsite problems and real-time change orders. The design-build process that allows for accelerated construction and real-time cost control requires contractors to flexibly modify every aspect of the construction project.

From the permitting process through the potential impacts of site conditions on the delivery timeline, and ultimately to final tenant change orders, every alteration in materials and design affects costs and delivery. For example, the possible discovery of site contaminants and municipality postponement of inspections by three to six months are not uncommon, and mishandling of sensitive information can potentially halt construction and can even cause the loss of tenants. The tools used in the field must be proven and reliable, and the data must be secure from the point of entry in the field to their use by all members of the development team. It also applies to content captured on mobile devices.

For gathering site photos or scanning paper documents with a mobile device, there are hundreds of free or inexpensive apps offered through iTunes and Google Play with the functionality to “scan” paper documents and convert to PDF. However, many scanning apps put sensitive data at risk by paying shockingly little attention to data protection and possibly even nefariously retaining and using that data in manners that would place intellectual property and trade secrets at risk. In the construction industry, the scanned content could be architectural drawings, contracts, service agreements, site plans, invoices or receipts (i.e., very sensitive information that should be handled with care to ensure it does not get leaked or compromised), or used after the fact in unintended ways (such as evidence). These risks extend to executives, employees and contractors and should not be taken lightly.

Consumer-grade document scanning apps are often free or $.99 and on the surface seem very competent at helping accomplish the simple task of using a mobile device to turn a paper document into a digital PDF. Some of the questions that should dictate the company’s choice on which apps are allowed and which are to be avoided are:

  • how and where is the data being stored;
  • what is the data used for;
  • how secure and private is the data; and
  • what is the app developers’ policy related to marketing (reselling) customer data?
It’s critical to understand that just as with industrial-grade construction materials, secure and industrial-grade apps are not simple to develop. Full-feature document scanning apps could take more than 500 engineering hours, plus the expense of branding, marketing, maintaining and hosting. The revenue model for most of these applications is not the initial $.99 or the fractional pennies earned when a user taps on a banner advertisement. More and more frequently, the business model involves the collection and distribution of user data, often in ways that would make the user feel extremely uncomfortable and could even conceivably pose liability issues long after a major project reaches completion.

Many apps offer optical character recognition (OCR) as a feature, which extracts text from documents. This is a very helpful feature when scanning business cards in order to convert them into contacts or to search the text of a document. There are two ways for an app to perform OCR: on the device itself, or by sending the file to a cloud server for processing and then passing the text file back to the device. OCR on a device is extremely rare because it:

  • balloons the size of the app;
  • requires a great deal of memory;
  • is harder for the developer to update; and
  • is generally less accurate.
Therefore, the wide majority of apps send scanned files to a cloud server in order to extract the text and send it back. Is that data transfer secure? Does the server retain a copy of the file or the extracted text? Does the developer sell or use this data in any way?

Another risk is unsecured cloud storage. Many apps will store copies of documents in the cloud for the benefit of allowing the user to access them from another device or a browser. How likely is it that the purveyors of such “free” storage are investing in encryption, multi-factor authentication, SSL tunneling, etc. to protect the user’s documents? Highly unlikely, which is fine for students needing free services, but unfortunate for construction executives seeking to avoid, rather than create, liability.

If an organization is scanning documents that contain business-critical information, it’s important to take a second look at the mobile document scanning apps that employees, executives and contractors are utilizing. The good news is that some apps actually do invest in security and data protection, as long as an organization is willing to pay a little bit more for the solution. The first thing to look for is an app that integrates with an Enterprise Mobility Management (EMM) platform. Organizations using MobileIron, AirWatch or Blackberry to manage devices will be well-served to find an application that has integrated with their data containers and passed their ISV security and penetration tests. It’s also worthwhile to look for more granular features, such as support for specific PDF versioning, like PDF/A or PDF 1.7, which lock the file down from future editing. Especially in construction, the ability to annotate documents with labels, drawings, arrows, etc. is useful. Most importantly, it’s critical to read the developer’s privacy policy to make sure there is a comfort level with their stated views on protecting information and that the company is not placed at risk for compliance and confidentiality clauses.

Basic steps to document scanning on a mobile device

  • Using the camera, a photo is taken of each page.
  • A complex method predicts the outer edges of the page.
  • The area outside the edges is cropped to create a polygon.
  • The polygon is normalized into a true rectangle.
  • The image is enhanced to remove “noise” so that the text is clearly distinguishable.
  • A filter is applied to remove all color from the image.
  • The images are converted to a PDF.

 Comments ({{Comments.length}})

  • {{comment.Name}}


    {{comment.DateCreated.slice(6, -2) | date: 'MMM d, y h:mm:ss a'}}

Leave a comment

Required! Not valid email!