I'm not a big gambler, but I think I understand the basic strategy of certain games. Take blackjack. The strategy is to keep track of the cards that have gone out. As the game progresses, you can better calculate your level of risk with each hand played. Once you have that understanding, you start betting more often and in higher amounts, based on your card observations.
It's the same in product development. Conducting design research up front reduces your risk of problems occurring downstream. As you move through the R&D process, you can invest more money more confidently. The betting here is your monetary investment in development, engineering, and prototyping. As your risk goes down, you can spend more money.
People sometimes have an inverse approach. They believe they have a great idea, and they throw a lot of time, money, and resources at it. Some even leap right into a concept, a prototype, and a working model. They've already spent a ton of money, but then feedback starts coming in, and changes need to be made -- some of them drastic. It takes time to make changes. At this point, the risk is really high, because of money already invested, and now there is a lot to lose.
It's a Catch-22. The point of design research is to provide the right information up front, so you can develop the best ideas early on. But nobody wants to spend the money to do it. In the words of Frank Lloyd Wright : "You can use an eraser on the drafting table or a sledge hammer on the construction site." The sledge hammer is a bit more expensive.
Design research minimizes your risk if you approach it correctly. What you're ultimately trying to do with research is understand what a user needs to do to be successful with the device. You might spend a small or moderate amount up front to get this done, but it will take you more time and more money to change it later.
You should walk away from this type of research thinking: "I have a checklist of everything this user group needs to do to be successful using this device to perform the job." Now your task is set before you, and that task is to design something that can check off every item on that list. When thinking this way, you don't end up designing what the customer requests. You actually design something that meets the needs of the target users that they have expressed in order to be successful using the device.
It's important to mention what design research is not: "Here's Option A and Option B. Which one do you like better?" This is just a simple focus-group comparison. Unfortunately, much development research is done this way, because so many people's first idea of research is "Let's make some prototypes, get them in front of users, and see what they say." It's wonderful to do that at some point, but not early in the process.
Finally, some people claim that doing design research puts