How to Ensure your Computer Vision Product is a Success
The power of computer vision is providing designers and business leaders new opportunities to solve our most challenging problems to deliver new product innovations. With that power comes increased complexity and increased risks of making poor decisions. In today’s business climate there is little room for products that miss the mark. We believe that focusing on the user experience (ux) is the key to unlocking great decision making and delivering solutions people truly need. At the Embedded Vision Summit, I had the opportunity to speak about reducing risk in computer vision design and increasing your success rate, by focusing on user experience design. This approach is not limited to vision-based products and can be applied to any new product or service, because in order to successful, every product must be desirable to the end user.
One very difficult risk in computer vision design is developing a product that is undesirable and/or unusable. This often happens when developers have a powerful technology and go looking for a solution which often leads to solving insignificant problems. As developers and business leaders we cannot afford to miss the mark here with such a large up-front investment. Without the significant research from stakeholders on what value the product could bring, developing the wrong product solution is very possible. Another risk is over designing and under designing a product. With this promising technology it’s easy as developers to get excited about new features and capabilities sometimes these are not desirable to the end user. On the other hand, you can also risk creating a product that is under designed and falls short on delivering critical value.
Google glass is a great example of a computer vision product with a plethora of dynamic choices on the path to success. From one perspective, Google glass represents one of the greatest innovations in the last decade. Packing a high-resolution display and camera, voice recognition, cap touch navigation, Bluetooth and WiFi connectivity all into one small wearable device was no small feat. It had, and still has great potential. From another perspective, it fell woefully short with its limited functionality, technology limitations, expensive cost and an awkward user experience. Google glass, like many failed projects, missed the mark for a combination of three high level reasons: First, pricing was not competitive, even as a development system. Second, the supporting technology was close but, not capable enough – battery didn’t last long enough, volume was too low, limited features etc. Lastly, and most importantly, users didn’t find the experience compelling, in fact they often found it disturbing. In summary, Google Glass failed because they failed to focus on the user experience.
So how do we reduce these types of risks that many of us are challenged with as developers? By focusing on user experience design. Within the user experience lie the needs and problems computer vision aims to solve and is the source of new innovations. Users are the ones that set the price for solving their needs and they determine how much they will pay for having certain needs met. Technology solutions must also work within the users’ context, capabilities and environment. If we know what the users want, what they are willing to pay for it, and what their competency is with technology, we will reduce the design risks and increase our success rate – guaranteed.
This is one piece to a valuable approach we call human-centered design (HCD), a framework for problem solving that keeps the user at the heart of design efforts. In addition to focusing on the users, there are three other core tenants, starting with the importance of balancing the three lenses of user desirability, technical feasibility and business viability. Desirability considers the stakeholder’s expressed and unexpressed needs by asking the question, “what does the user need?” Feasibility considers organizational and technology capabilities to stated expectations by asking “are the needed capabilities feasible?” And expected revenue, profit and cost models are considered with business viability by asking “is there a viable business?” As engineers, it is easy to naturally gravitate to the feasibility domain and forget about the user who is the one experiencing our designs. As Steve Jobs aptly put it, “You’ve got to start with the customer experience and work back toward the technology.” The third of these four core tenants is experiment and iterate. HCD is agile, highly iterative and lean, and these lean methods are what keep iteration time and costs to a minimum. The fourth is embrace collaboration. More collaboration leads to the best solutions as one idea begets another. These four attitudes need to prevail throughout HCD in order to improve the likelihood of success.
The HCD framework has a collection of tools that align with a series of development stages that are helpful before and during the development process. These four distinct stages alternately diverge and converge as the number of elements in play expands and converges. Each stage has an applicable set of collaborative tools for driving progress and helps to keep the HCD mindset at the forefront of your thinking. Discover focuses on a deep look at stakeholder needs and experiences, market trends and industry dynamics. This is where we get close to the users and gather insights that serves as the foundation for the rest of the effort. Analyze focuses on fully understanding user needs and distilling them into key success factors to fuel the creative process. In my experience, these key success factors represent the real power of HCD. Acting as guardrails, they drive ever increasing focus and direction in follow on stages. Create focuses on generating potential concepts and solutions, evaluating them against the key success factors. When you reach the create phase, you are able to more successfully apply ux design to the new concept, due to the extensive research done upfront. And lastly, Develop focuses on clarifying solution(s) and validating them with key stakeholders. The feedback gathered through this process guides the refinement of the solution into the best possible solution to be advanced. If you’re interested in learning more about these stages, you can view an outline of this process here.
At Twisthink, we have been navigating the challenges of ambiguity on a daily basis, for nearly two decades. Doing this well is critical to the success of any product or business. If you are interested in discussing how you can apply this human-centered design process to your organization, let us know! Or, if you are in need of a partner to come alongside you to work through these challenges and risks, in order to improve ux optimization, we would be happy to help.
A fictitious story (for now) with real implications on how Embedded AI unlocks the value of IOT. Whenever I hear the term “the edge” in reference to artificial intelligence (AI) on the edge, I can’t help but think of Dave Evans – the Rock & Roll Hall of Fame...read more
Most people can agree that our machines are becoming faster, better, and, ideally, cheaper. But more intelligent? That’s where some people draw the line and many outside of technology spheres get a bit nervous. Well, here’s your spoiler alert: you are already highly...read more
Artificial intelligence (AI) is all the buzz right now and many businesses are looking for ways to flex their technology and innovation muscles by trying to find a way to implement AI. While AI can provide your business with an edge when applied in the right ways,...read more