We’ve all heard about the recent Apple Maps fiasco. The app misplaces airports, loses buildings, gives directions that send users straight into the ocean, and omits entire towns. What went wrong at Apple? What lessons can we learn from this to make sure that your connected product doesn’t become the next Apple Maps? In a word, verification.
We have observed that great products are often victories of verification. Killer connected products come about when you invest the time and resources into verifying that your product’s “minimum value proposition” works flawlessly. Apple Maps failed because Apple’s culture of secrecy prevented developers from rigorously testing and verifying the core features of the app. The fiasco even costed one senior executive his job. Let’s look at another example which demonstrates the payoff of proper verification.
Consider the story of STANLEY. In 2005, the Stanford Artificial Intelligence Laboratory Racing Team created an autonomous robotic car, STANLEY, which went on to win the 2005 DARPA Grand Challenge, a 132-mile driverless car competition. In the previous year’s challenge, the best self-driving vehicle had only managed a measly seven miles before it hit a berm and caught fire. Why did STANLEY succeed when so many others failed?
Six months before the competition, the Stanford Racing Team froze all software and hardware development and dedicated all its resources to debugging the core STANLEY features. It says a lot when some of the greatest minds in computer science devote half of their project schedule to verification.
When resources are limited, verifying that the core features of your product are bulletproof is always better than adding new, non-critical features. Take the first iPhone for example. The first iPhone had many nice secondary features, yet it couldn’t multitask (e.g. read an email while the GPS was running) which was a feature that most customers relied on and expected from their PC experiences.
Product verification is one of Arrayent’s core values, and we’re convinced that it is the second key to a killer connected product (the first key, as we discussed in our last blog, is low cost). Verification is often the difference between a catastrophic flop like Apple Maps, and a monumental success like STANLEY.
By Kayce Basques