Pictured Above: Rick Lorenzen holding a ruggedized laptop showing data from a Mille Lacs lake survey.

The impact that technology has on the everyday lives of Minnesotans is changing exponentially. As we head into 2020, we wanted to reflect on where we were 10 years ago, the challenges and opportunities addressed by technology today, and where we might be headed in 2030. Technology touches nearly every aspect of the services that Minnesota state government provides, and the evolution of those services is underpinned by advances in technology, better business processes propelled by new applications or data management, and a focus on access for all.

Over the last few weeks, we have been sharing the perspectives of employees from Minnesota IT Services (MNIT) who work hard every day for the State to ensure that Minnesotans have access to a better government.

Rick Lorenzen is an IT supervisor for the Ecological and Water Resources Office where Minnesota IT Services (MNIT) partners with the Minnesota Department of Natural Resources (DNR). He says that one of the biggest changes he’s seen with technology over the past decade is the constant feeling of being “online”. He takes advantage of “airplane mode” when he’s able to be more intentional about his time.

Q&A with Rick Lorenzen

What were some technology issues that you faced in 2010 that have been addressed as we start 2020?

Rick: In 2010, the lack of high-speed connectivity and ability to store information digitally were issues. The DNR has many outstate locations that either had terrible bandwidth or didn’t have wireless connections available. Even in our central office in St. Paul, we would have trouble viewing video because of bandwidth issues.

While data storage issues have mostly been resolved as technology advanced, some of our issues connecting to remote or off-line sites have remained an issue. At the beginning of the past decade, however, we developed a first of its kind system to collect data offline and store it so it could be accessed and loaded to a database at a later point. The system was used to record fish surveys for DNR staff, replacing waterproof paper and pencils. Now we have many offline data collection applications that allow staff to be in the field, or on the water collecting data for later analysis.

How has technology changed the way that the DNR operates as an agency?

Rick: Since I began programming in 1980, business management relied on technologists to manage all the data and their eyes would glaze over during technical conversations. Now, management at DNR, and other organizations, has a lot of interest in the data. A few years ago, it began with an interest in discussing data as a general term – what do we have, how do we use it, how do we protect it. Now, these conversations are shifting into the actual needs of the business – our agency partners. We are finding ourselves responsible for making the data accessible and storing it, and our partners at DNR are making critical business decisions about what to do with it. That is the difference: technical data stewardship versus data ownership. This dynamic has created a stronger relationship with our agency partners at the DNR, we are brought in as technologists to provide tools to help analyze the data for the DNR as a business, and ultimately have an innovative influence on how the data will serve all Minnesotans.

What are some changes in data collection that you expect to see by 2030?

Rick: I’m hoping we see further advancement in data integration. We have literally hundreds, if not thousands, of small data collections by biologists and researchers that have fulfilled their original purpose for collecting that data for a specific study. However, a large percentage of these data collections, do not contain a unique waterbody identifier because the data collector knew which waterbodies were involved in the study and used an abbreviated name field to tell themselves which lakes were involved during their analysis to produce the report. It’s the same for many other collected data attributes –researchers use abbreviated species names instead of a unique identification code. Without these unique identification codes, we cannot easily connect to worldwide data collections to gain the advantages provided by “big data”. We are now designing systems that ensure a unique identifier is included as part of the data being collected by providing drop-down selection lists for waterbodies and species. Wherever data is being collected, we need to use big data identifiers to match it up with data from other states, federal data, and around the world.

Another hope is that researchers will rely more on analyzing data from sensors and point clouds than they do on data that was hand-collected by humans. Up to now, we have relied more on our human researchers and interns to collect data in the field, but they are human and have the potential to make mistakes. In the beginning of the past decade, some researchers insisted they needed to be able to enter location coordinates by hand because some of their GPS hardware didn’t easily connect with their computer applications. Some of the typed-in locations that were supposed to be on a lake in Minnesota, showed up in Canada or South America when displayed on a map because of human error. Now, we can take for granted that the device used to record the survey results will record the location accurately and no one is retyping location coordinates anymore.

We do work on our end as technologists to minimize human errors during data collection. We can ensure the length of a fish entered is valid for the specific kind of fish while the fish is still in hand. Unfortunately, we cannot prevent some errors, like when a human selects the wrong species from a drop-down list of valid species. Researchers know that when we rely on humans to input data into systems, some portion of the data contains erroneous “noise” that must be filtered out by data analysts.

Sensors could help with this and help us to identify things in a less error-prone environment. For example, there are researchers using pictures interpreted by special software to identify and track monster-size whale sharks by their markings instead of by embedding tags in their flesh. The special software was adapted from software created for astronomers to identify star clusters in space. Before 2030, I hope we will be able to survey our fish and even identify and track a unique lake trout by taking its picture instead of catching it in a high-mortality gill net.