RELATIONAL MANAGEMENT and DISPLAY of SITE ENVIRONMENTAL DATA - PART 6 ppsx

It is a very frustrating feeling to work hard to get a database set up, overcome the learning curve, perhaps put a lot of effort into entering and/or formatting data, and then find out t

PART SIX - PROBLEMS, BENEFITS, AND SUCCESSES

CHAPTER 25

AVOIDING PROBLEMS

Implementing an EDMS can be either a positive or a negative experience, depending on how the project goes When it goes well, it can be very rewarding When it goes poorly, it can be very frustrating (at best) More often than not, the result is positive, and there are specific things that can

be done to increase the chance of a positive outcome If you really think through the issues that were discussed in Part Two of this book, you will be well along to anticipating and avoiding problems This chapter discusses some of the possible pitfalls in implementing an EDMS

MANAGE EXPECTATIONS

A very important aspect of implementing an EDMS is to make sure that the people who are designing the solution and those who will be using it are talking to each other This is true whether the solution is being bought or built, and it goes both ways The implementers of the system must

be aware of the needs of the users so that they can satisfy as many of those needs as possible The users must be aware of what is actually being implemented so they are not surprised when it arrives

on their desk

Understand the real needs

If you ask people what tools they need to do their work, the results can sometimes be surprising One time a senior vice president of the environmental group at a major company was asked what would make his job easier He didn’t hesitate for a second before he replied

“retirement.” Obviously nothing that could be done with the software would help with that When you are trying to define functionality, be sure to focus on tools and processes that will make a significant contribution to improving the job performance of the people using the system Statements that start with “It would be cool if …” should be examined critically Sometimes they contain real needs Sometimes they don’t And sometimes it’s hard to tell the difference On the other hand, a statement that starts with “My regulators require …” probably contains an important requirement in order for the EDMS to be successful

Don’t promise the world

When you’re in the early stages of implementing an EDMS (or any system for that matter) it is easy for the developer to respond to every request with “Yeah, it’ll do that.” Those statements

usually come back to haunt you Be sure that when you say the software will do something, you really understand what it involves to provide the functionality being requested, how that functionality fits with the design plan, and whether resources can be provided to fit that need Also, don’t forget that with software users (as with kids) “no” means “maybe” and “maybe” means “yes.”

Plan adequate time and budget

The dismal success record of software implementation projects was discussed in Chapter 8, as were some thoughts on estimating a budget for an implementation project Nearly any project can

be completed given enough time and money The trick is for the developer to implement the greatest amount of the most useful functionality within the schedule and budget available Experience has shown that nearly everything takes longer and costs more than was originally anticipated Experienced project managers have a habit of doubling everything If this is factored into the planning process, rather than dealt with as it comes up, success is much more likely

Manage risk

Given that the project has a finite, and in fact relatively large, chance of failure, it is important

to plan for the risks that might derail the project You should ask questions like: What happens if a key developer becomes unavailable? What if management changes during the project? What happens if the budget is cut part way through the project? Or, in the worst case, what happens if the project fails completely? Good project managers have the ability to react to change, and can often salvage something, or even pull the whole thing off, when things go poorly Be sure to have a backup plan for the most likely contingencies Any but the smallest projects should have a formal Risk Management Plan This type of thinking is uncomfortable for many people, but really does pay off when things don’t go as planned

USE THE RIGHT TOOL

There is a tendency to use the tools that you have and are familiar with This is particularly true with software If people know how to use a spreadsheet or word processor, they will want to use that tool to manage their data, even if it is not appropriate And even if you choose the right tool, the size you choose must fit the task to be accomplished

Avoid overkill

At first glance it would seem that choosing a more powerful database system than you need might not be too bad It seems that you can just use what you need, and not worry about the rest This is not always the case Overkill in database software hurts you in several ways First, it can require the expenditure of resources beyond what is necessary, putting a strain on other parts of the organization or project Second, big software can have a big learning curve, and it may be a challenge to get people up to speed on using it Finally, and this is the worst, if the software is too big and complicated, people won’t use it, and will go back to their spreadsheets

A common case of overkill in database projects is in choosing the data repository There is a tendency to use a big server database system even for small databases just because it is there, or because “It is the company standard.” Don’t fall into this trap Using too big a hammer isn’t good for the nail, or for the fingers holding it If the database will be small, be smart and use the appropriately sized software to manage it

Also avoid underkill

The problems of underkill are more obvious, but no less serious, than those of overkill It is a very frustrating feeling to work hard to get a database set up, overcome the learning curve, perhaps put a lot of effort into entering and/or formatting data, and then find out that the software won’t do something that you need Often the time spent organizing data will not be lost if you have to make

a transition to a more powerful system, but you will probably have to repeat the time (and cost) to get up to speed on the new system Sometimes the path is not too hard, such as in moving from Access to SQL Server, but other times it can be painful It’s better to plan ahead and take your long-term needs into consideration before you select software

Buy vs build

This issue was discussed in Chapter 8, but is worth bringing up again here There is a tendency

to feel that either your project is unique or too complicated, or your skills are so high, that the only answer is a custom solution Sometimes this is the case, and sometimes it is not Usually, it is much less expensive to buy rather than to build, as long as what you get fits your needs Put your personal preferences aside and think hard about what is best for the project, given time and budget limitations and technical requirements, and make your decision based on those factors

PREPARE FOR PROBLEMS WITH THE DATA

Getting the software up and running is the easy part Finding, organizing, and entering data, and keeping it flowing on an ongoing basis, is the bigger challenge There will almost always be problems with the data, ranging from minor to severe It takes skill, patience, and perseverance, and often help from others, to be successful

Where is the data?

The first problem is finding the data to begin with Whether you are dealing with hard copy or digital data, locating disks or documents can be a challenge Particularly on projects where the personnel working on the project has changed over time, even finding someone who knows what data there is (or was) can be difficult The amount of effort spent on locating and organizing data should be commensurate with the value of having the data in digital form

Structure problems

Once you have found the data, you will need to find out how it is organized With hard copy data this usually isn’t too hard because you can look at it and (hopefully) figure it out, but if people are entering the data, changes to formats over time can be a problem With digital data it can be even worse, since it’s not unusual for the format of digital data, even from the same source, to change often It takes a patient person who is knowledgeable about the data to figure out and accommodate changes to the data structure prior to importing it

Content problems

Once you know the structure of the reports or files, you must be sure that you understand the data they contain Different data creators may have different ideas about how to report data For example, when a chemical value is less than the instrument detection limit, it can be reported as

zero, the detection limit, or half the detection limit If the detection limit is included in the file, you can figure it out, but it is important to really look closely at your data, and be sure you understand what is there, before you try to use it

Delivery problems

When you are working with data on an ongoing basis, you want to make sure that the process you have in place for getting the data in a timely fashion is working well Often the time between a sampling event and the due date for a report can be short enough that any problems in data delivery, import, checking, or reporting can stress the deadline Be sure to plan ahead and know where your data is and when reports are due so you can minimize timing problems

PLAN PROJECT ADMINISTRATION

The final pitfall to be avoided is inadequate planning for the administration of the project You should determine up-front who will be in charge of the management of the project, the personnel responsible for implementing the project, and then for running the system after it is operational Interfaces with outside organizations, such as labs, the IT department, and any other affected personnel should be mapped out and discussed in detail at the start Finally, the project must be supported by a realistic schedule and an adequate budget

INCREASING THE CHANCE OF A POSITIVE OUTCOME

If we were to boil this section down to one thing, it would be communication If you involve

all of the people affected by the system from the beginning, and keep them involved throughout the process, the chance of success is greatly increased Plan and design thoroughly, so the software developer or vendor and the users are always on the same page, and keep the communication up during implementation Also, be sure to take advantage of peripheral benefits like improved communication that can occur in other areas once people start talking about the database

If you do all of the above things well, your EDMS implementation project has an excellent chance of success, and you can reap the benefits of a job well done Implementing a database system can have a tremendous positive impact on a project and an organization It’s worth the effort on everyone’s part to make it be successful

CHAPTER 26

SUCCESS STORIES

Don’t let the previous chapter scare you Most data management systems, if implemented carefully, are judged a success by those using them, once they are up and running This section will provide some examples of the benefits of an organized data management system, as recognized by project management personnel In the following examples, the problems were successfully overcome (or even better, anticipated and avoided) so that everyone was satisfied with the outcome

FINANCIAL BENEFITS

Feng and Easley (1999) provide some examples of benefits that they saw on an environmental management information system they implemented These included upgrading their system to get off their mainframe, reducing costs by streamlining workflow and reducing compliance costs, improving compliance by handling their data better, and anticipating the future by preparing for future regulations Often the justification for implementing a data management system is primarily financial Financial benefits that can be expected by implementing a data management system fall into several areas These include reduced overhead costs, increased project efficiency, replacing older systems, reduced project operating costs, and increased revenue

Reduced overhead costs

Lower costs can be achieved both on the data management component of projects, as well as

by using the data management system to improve other areas of the project A good example of decreasing overhead costs occurs when the data management work can be transferred to a less expensive employee after implementation of an easy-to-use data management system For example, one company was able to transfer much of the data management activities for a complex project from a high priced project manager to more economic tech and clerical staff members This resulted in average savings of $25 per hour on about 40 hours per month, resulting in savings of

$12,000 per year on this one project alone

Increased efficiency

One of the most obvious areas of financial benefit of an automation project is increased efficiency One Enviro Data user reported that its time to process electronic deliverables from its laboratories decreased from 30 minutes to 5 minutes per file after it implemented and enforced a data transfer standard and a closed-loop reference file system This implementation helped the laboratory to deliver clean data Since the data administrator was handling about 300 files a year, this translates to 125 hours per year saved, for cost savings of almost $5,000 per year just for that one task Additional savings were realized in increased efficiency in selecting and reporting data

Replacing expensive older systems

Financial benefits from replacing expensive in-house systems (sometimes called “legacy systems”) can result in both software and hardware savings The data validation manager for a national consulting company was using the company’s in-house database that could no longer be supported, but still needed to perform the same functions The DOS-based database was old and too difficult and expensive to maintain The company needed a commercial system that would allow it to do data validation without the cost of maintaining the old system The company purchased an off-the-shelf solution that was then customized slightly for its specific project needs The company now has a better tool for doing its work for less money than maintaining the old system

Very significant cost savings can be realized when the data management system can be moved from expensive mainframe hardware to less expensive personal computers and servers Maintaining a mainframe can cost several thousand dollars a month, compared to several hundred dollars a month to maintain a PC-based database server It doesn’t take too many months of this to pay for a system upgrade, and then after that the cost savings go straight to the bottom line

Reducing project operating costs

An EDMS can help lower project operating costs in areas other than data management One large industrial company with many facilities routinely uses its EDMS to review groundwater monitoring wells to identify ones where concentrations are consistently below regulatory limits With a database of several thousand wells, the company is able to identify at least two wells per quarter that can safely be monitored less often Each well that can be sampled annually instead of quarterly saves about $3,000, and the database provides the documentation to take its case to the regulators If the company is successful on half of the requests, it can save $12,000 per year for the four wells, and these savings are cumulative from year to year

Increased revenue

An environmental engineering company needed to provide a data management system for one

of its clients, a pipeline company The client needed to manage its historical and ongoing data in a standardized way because it was being pressured by its regulators (EPA) to provide more comprehensive reporting of the trends in contamination levels The engineering company wanted

to make money by solving the client’s problem, and to generate a satisfied customer who would come back for more services By showing its proficiency in using an efficient EDMS, the company landed a $300,000 data management task Only 20% of that actually went into data management costs, resulting in increased revenue of $240,000 for that project

Taking the financial savings from three of the examples above and adding them up, the total is

$27,000 per year This means that every month that the implementation of a data management system is delayed, $2,250 is lost Another way of looking at it is that if implementing the system takes $75,000 for software, training, data conversion, etc., then the time to pay out the investment

is 33 months, for a return on investment (annualized IRR) of 23% over 5 years This would be considered a good use of funds in most organizations To this can be added the technical and intangible benefits described in the next two sections

TECHNICAL BENEFITS

While the dollars usually drive the purchase decision, the technical benefits are often the greatest benefit from an EDMS implementation project Building a comprehensive, centralized, open database can generate improved technical results in a variety of ways The biggest technical benefit is the improved quality that results from removal of database fragmentation People are always using the best data available, not an outdated data set, or one that was thrown together to answer one question, and then used later to answer a different question, which might really have different data requirements Related to this is improved communication on the project, because everyone is looking at the same data This results in increased confidence in the data and in the decision-making process for the project

The impact of these technical benefits on those outside the project, such as upper management and especially regulators, can be significant If these people develop confidence that the project team is staying on top of issues at the site, the result can be less scrutiny, and consequently less aggravation, for the project team If they are finding and reliably dealing with issues as they come

up, the project goes more smoothly for everyone

Increased efficiency was discussed above from a financial perspective, but can also be viewed from a technical perspective If the project team spends less time on repetitive activities like cleaning up data and moving it around, the members will have more time to spend really working with the data, which will result in a better understanding of the site and better management of site issues

Real access to data

People like to point out that the difference between data and information is whether you can use it or not A geologist that Geotech worked with had a large amount of base map and well data that he had obtained from a mainframe system when the management of the project changed, and the mainframe became unavailable He had the well data in thick printed reports and in digital files from a mainframe For a while he worked with the paper copies, but became frustrated with how hard it was to find things Even though he was not very computer literate, and had never used a database before, he knew exactly what he wanted, which was to be able to select the data in specific ways, and to post the results on a map Based on his very specific instructions, it was easy

to build a data management system to satisfy his specific needs The mainframe data was imported and cleaned up, the base map data was reformatted for his mapping program, and he was trained on how to select data and place it on a map His frustration went away, because now he could easily use the large amount of data that he had

A mining company had a problem with a relatively small data set, but with very complicated retrieval requirements It had to create an annual report that was taking them many weeks of

One environmental lawyer in a small town can starve to death, but two can make a good living

Rich (1996)

counting and re-counting the data, with various different ways of selecting and grouping the data The data included public health data, and in addition to patient confidentiality issues, its regulator was requiring that decisions be made on relatively small data sets, so accuracy in retrieval was very important Building a password-protected system to hold the data was relatively easy Automating the retrieval was more complicated, and ended up requiring 267 SQL queries nested up to six deep with multiple unions to create the 17 output tables that made up the annual report Now the company can create all of the data for the report in just a few minutes each year when it is due, and has confidence that it has been done right

There are many benefits that can result from better access to the data More and more, people with a legitimate need for access to data can be provided this access efficiently and cost-effectively using client-server and Web-based tools Having the data in a centralized, open database is the key

to this

Better visualization

Another technical benefit is the ability to analyze the project better Having a comprehensive database opens the door for better analysis and visualization tools, which can lead to a better understanding of the project, and a better ability to anticipate and minimize problems before they become critical This results in a process where projects are managed by the project team, rather than the projects managing the team with a series of crises and fire drills

An example of benefits from better data management and visualization involved public health data The client was gathering exposure data, and needed to organize it and try to identify the source of the exposure Geotech implemented a relational data management system to help the client store and check the data This system integrated the capability to draw the data on a graph as

a time sequence and on a map as a function of distance from the suspected source The displays that were created provided useful insights into the mechanisms leading to exposure and impact so the exposure could be minimized

SUBJECTIVE BENEFITS

Some benefits derived from improved data management are intangible, but still contribute significantly to the overall success of the project Data management can be the most tedious component of a project Implementing an efficient system for processing data can significantly improve morale, which results in improved quality of output, less staff dissatisfaction and turnover, and in general a happier and more productive project team

Minimized drudgery

Sometimes the database software takes on work that it is hard to get people to do One organization had several projects where the routine drafting was verging on drudgery Project personnel were printing out their data from spreadsheets, and then the drafters were typing it into their CAD program In another case, they were printing out results and pasting them on drafted maps By moving the data into a database that allowed sorting and gave good control over creating formatted output, and by integrating a GIS to automate the display of the data, they were able to automate the process just in time to prevent mutiny

Improved communication

A national landfill company client had a problem with where its data was going to live It had

laboratory and other data on an ongoing basis for all of these facilities It wanted its hydrologists to

be able to call up data on any of the facilities any time it was necessary, and then be able to work

on the data locally, even on an airplane, with some confidence that it is up to date The solution was a software system that was distributed between the main office and each user’s laptop Clerical staff in the main office imported and checked the laboratory data as it arrived Remote users downloaded subsets of the data for local use, and then discarded them when they were done with them As long as they were working with a relatively small amount of data, they were very happy with this process If they needed to work with larger sets of data, they either needed a fast connection, or to have a subset made and sent to them Either way, they were able to efficiently work with the data they needed

In another example, a project required the combination of environmental data with health data The environmental data was gathered and entered in one location, and the health data in another Patient confidentiality made it very difficult to work between the locations, because people on the environmental side could not see all of the data on the health side, and the health people were not entering the environmental data The problem was that they were dependent on each other, with the health people needing property information from the environmental folks, and the environmental folks needing family and other information from the health side A system was built that allowed data to be entered at either location, and then transferred, as much as allowed by confidentiality, between the locations It also allowed the creation of digital output combining the two types of data while still maintaining confidentiality, so that toxicological and other studies could be performed with the greatest amount of data

Increased confidence

A very successful database implementation was for a medium-sized environmental company The system that was implemented helped the company significantly throughout the entire life cycle

of the data from field data gathering to submission of results to regulators The client had several projects where it needed to gather data in the field and from laboratories, import and edit it, create printed reports, and deliver digital data to the regulators A system was put in place that could be used on laptops in the field as well as in the office The software helped the client enter and check historical data from hard copy, import data on an ongoing basis from the laboratory, and perform statistics and create reports Then the software was sent along with the data to the regulators, so they could work with the data themselves Both the employees of the environmental company and the regulators worked with the data so intensely that they felt they had worked out all of the problems with it, and that the data in the database could be trusted This improved confidence and trust between all parties, from which everyone benefited

In another example, the senior vice president of environmental affairs for a major oil company had numerous consultants managing the company’s data, and the oil company was losing data because of consultant turnover The data was managed remotely and the company had no control over it The company was paying for data but not getting it It needed a centralized database so once it paid for analyses, it would always have them A centralized in-house database was implemented, and the company now has complete control of its data That database now has over a million records of analytical data in it and is growing daily