The story of Samuel Pierpont Langley is fascinating, primarily because most people have never heard of him. Langley was a well-regarded scientist in the late 19th century with extraordinary qualifications, including being the Secretary of the Smithsonian Institution. Langley was obsessed with taking his place in science history and decided the way he would achieve this is by creating the world’s first manned flying machine.

Langley called this flying machine “the Aerodrome,” and with access to large amounts of capital and powerful friends, he set about on development of his project. As Langley started his research, he became focused on the single problem of aircraft propulsion, noting that almost anything would fly if one could get it moving fast enough. So, he invested most of his time and over half of the investment money into a houseboat with a catapult from which he would launch his Aerodrome.

Five years later, when the Aerodrome was finally ready, Langley took his device to the middle of the Potomac river outside of Washington D.C., and under all sorts of fanfare and media attention, he launched the Aerodrome. Immediately, the Aerodrome’s wing clipped the catapult and went straight into the river. It “flew like mortar,” according to a journalist at the time. A few months later, he tried again, and this time the Aerodrome and catapult both collapsed through the houseboat before launch. That was the end of Samuel Pierpont Langley’s efforts toward manned flight.

Data Science, broadly defined, is a discipline that has been around for many years. But, similar to Langley’s Aerodrome project, a frustratingly high number of big data and artificial intelligence projects still fail before take-off. In fact, according to Gartner, only 20% of data science deliver business value. Many of the reasons for project failures parallel the approach of Samuel Pierpont Langley.

Samuel Pierpont Langley made many mistakes in his efforts toward manned flight, but one mistake stands out from all others. As with many current data science and artificial intelligence projects, Langley’s failure can be traced back to starting with the wrong question and thereby framing the opportunity and the overall effort incorrectly. Langley became consumed with his research into propulsion at the expense of aircraft design and controlled flight.

All too often, data science projects begin by analyzing data with the expectation that an interesting insight will reveal itself and become the basis for the business case that justifies the effort. This approach can quickly waste corporate resources and result in valuable data scientists trying to produce insights that likely will not benefit the business.

The challenge of framing business problems correctly and ensuring a process is producing actionable results is one that has also been tackled with the creative problem-solving process called design thinking. Unfortunately, however, too many companies have separated “designers” and “data scientists” in-house and are not benefiting from the process improvements and disruptive ideas that can be fostered by bringing the two disciplines closer together.

The combination of these two skill sets is an approach to solving data science and artificial intelligence problems with ingredients such as a user-centered focus, bias toward action, and a culture of rapid prototyping. Design thinking allows data scientists to scope specific and meaningful challenges, and to synthesize findings into compelling insights. As MIT CISR research suggests, this also accelerates corporate learning and limits the risk of expensive failures. Once the challenge or hypothesis is clear, the design thinking process provides the opportunity to generate radical design alternatives, to prototype ideas and to continually refine the data science solutions to make them better.

At the same time Langley was buried deep in the study of propulsion, two brothers were at work on the same manned flight challenge, but were approaching it from a different angle. Their only investment came directly from the sales in their little Dayton, Ohio bike repair and rental shop. They didn’t have deep pockets or powerful friends, but they were great innovators with a knack for both science and design. Rather than being consumed with propulsion, the Wright brothers studied the flight of pigeons and focused their design on the human-centered ability to balance and control the aircraft. Orville and Wilbur Wright completed rounds of experimentation with kites and gliders in the years before their flight, prototyping, innovating and adjusting as they went. The aircraft that they ultimately flew at Kitty Hawk took flight just nine days after Langley’s spectacular failure and cost significantly less. Within a year, the Wright brothers were flying five miles and within two years they were flying 25 miles.

Data science and artificial intelligence are powerful forces that can be made even more powerful with creativity, innovation and a design mentality. Instead of approaching data science by narrowly focusing on the use of new algorithms or building better statistical models, a design-thinking approach recognizes data scientists as creative problem solvers. Engaging data scientists in design thinking can produce powerful insights and can unlock empathy for the people who will be touched by the data engines they develop. Ultimately, it can create an environment that will empower data scientists to develop solutions the business has never dreamed of. 

Learn more about how Avanade can help you creatively solve your data science and artificial intelligence challenges.