3 Tips for Avoiding the Perils of Dirty, Messy Data
3 Tips for Avoiding the Perils of Dirty, Messy Data
The market pressure for companies to invest in sustainability is growing, and solid data is the foundation for building systems that foster and support sustainability.
But what can companies do to ensure they secure good data?
GreenBiz Senior Contributor Paul Baier, who leads Groom Energy's sustainability consulting practice, Emma Stewart, Autodesk's senior manger for architecture, engineering and construction sustainability, and Jim Sullivan, vice president for sustainability management and strategy at SAP, offered their advice on the subject of "Turning Good Data into Smart Systems," during GreenBiz Group's VERGE conference this week.
Conversations about the need for systems thinking and a free flow of information among all elements of an organization threaded through the many discussions that made up the VERGE conference -- a series of livecast programs in based Shanghai, London and San Francisco that focused on the convergence of vehicle, energy, building and information technology and how their intersection will affect business sustainability.
"On this theme of 'Turning Good Data into Smart Systems,' I think as we transition to a knowledge economy and one where compute power and data storage are actually very cheap relative to human labor, we will increasingly encounter a number of data-related challenges," said Stewart in the panel discussion moderated by Baier.
Stewart identified five main challenges:
1. Messy data, which is not sanitized, consistent or "organized appropriately into hierarchies that can be used either by humans or computers," she said.
2.Too much data. Though plentiful, the sheer amount of data obscures its significance and the insights needed to make good decisions.
3. Non-intuitive data, which is not understood by non-expert stakeholders or requires significant effort to parse out. This can be a big problem with public infrastructure projects, which require political and community support in order to get underway.
4. Static data, which results in decision-making based on information that can be days or months old. In management of buildings and entire real estate portfolios, this problem can lead to mounting costs and wasted energy and other resources.
5. Absent data. This can be data that is simply not available or too expensive to collect.
Regarding the last item, Stewart said, "as the world economy clambers its way out of this global recession, we're making major investments in upgrading existing infrastructure. But to collect the data about that infrastructure has traditionally involved many, many weeks of onsite visits, often by very highly qualified, expensive engineers. That needs to be democratized and, ultimately, automated if we are to reach the scale needed for an economic and environmental recovery."
She suggested three steps for companies to take to ensure they get good data:
1. Make data clean and streamlined:
-- Clean it by reorganizing data into centralized repositories with matching hierarchies so it can be sent from one database to another.
-- Streamline it by prioritizing the data that is most relevant to decision-making.
2. Make the data intuitive:
-- Doing so makes the data a compelling tool for many types of users.
-- Ways to do that include using 3D presentations and overlaying complex datasets in single view that is information rich as well as easily understood.
3. Make the data smart, which can mean that:
-- Users can access real-time feedback with portable smart devices, like smartphones or pads.
-- Data on past performance and variables like the weather can be drawn upon to provide simulations of performance, so that users can predict behaviors of various things, such as energy consumption in a building, stress tolerance or recyclability of a product in manufacturing, or the way floodplains or rivers behave.
"If we as a society want to turn good data into smart systems," we really have to consider the five challenges and address the challenges by making data clean, streamlined and intuitive," said Stewart. "Then we have chance of making it smart."
Discussing the reasons why companies want and need data, Baier recapped five drivers for sustainability investment and reporting.
Some leading companies are now "treating their carbon disclosures like financial disclosures," he noted.
The five drivers Baier cited are:
1. Requests from top customers.
2. A desire to improve company brand or image.
3. Cost savings.
4. Investor pressure.
5. Greenhouse gas regulations and compliance.
Sullivan of SAP, a $18 billion firm whose business management software serves as the system of record for some of largest and most complex firms, outlined the business case for sustainability -- which further emphasized the need for good data, strong analytics and compelling presentations.
Key points of the business case include:
1. Strategy for sustaining for the business model, employability, health and diversity, and higher market cap, which can be supported by sustainability reporting and analytics.
2. Risk reduction as it pertains to the costs and risks of compliance and noncompliance and to customer and legal requirements, which all boil down to operational risk management.
3. Resource productivity, which translates to cost reduction through process efficiencies and supply chain collaboration and is achieved through energy and environmental resource management.
4. Sustainable consumption, which can enhance competitiveness, help build new markets, increase brand value and boost customer loyalty.
Sullivan also provided a series of front- and backend examples of how SAP is helping businesses and government agencies run better and provide transparency for their operations. He pointed to www.recovery.gov and experience.sap.com as resources for viewing SAP business intelligence software in action.
The webcast of the panel discussion and the other VERGE programs are being archived and will be available for three months. Access is free with registration.
Photo CC-licensed by David J Morgan.