Today I was asked to give some educated guesses about industries that were ripe for disruption. One immediately came to mind was automobiles. Lots of forces that everyone talks about, but a last year a young friend mentioned it was frustrating that she couldn't send a message to her car.
It is odd when you think about it. Most of us walk around with network connected pocket computers and think nothing of communicating with people around the world in many ways. These computers have a growing number of sensors that provide information to us and other people or systems. The most basic new cars have an order of magnitude more sensors than the most complex phones. These are connected to a few networks and usually dozens of small computers (usually called electronic control units) that mostly lied buried away making sure some rather sophisticated systems mostly function. But we are more aware of an infotainment layer - a layer that usually has a terrible user interface and user experience. Not only is the experience bad, but we are stuck with it for the life of the car.1
There are a lot of reasons why the industry moves slowly on automotive software. I won't get into that now as I only have an hour, but the infotainment systems do not require the type of careful design and testing required by drive train, stability control or restraint deployment systems. In theory these could be more like PCs or smartphones.
Or - as my friend suggests - perhaps they should be designed to use her iPhone as a network connected controller and display. Perhaps the car is properly just the transportation attachment to her phone and networked life.
Others have thought about that - Automatic uses a small box that connects to your car's OBD-II data link connector.2 This allows them to tap into realtime information that normally is only available to your mechanic and beams it to your Android or iOS phone via Bluetooth. Now a lot of clever things can be done on your phone.
What impresses me is not what it does as it could do much more - but that no car maker has gone this far. I'm guessing they have a half dozen competent software engineers. A large car maker has hundreds - but no imagination. This is basically a hack using an available port that was never intended to do this sort of thing, but it works and people will see it as amazing and desirable.
Much has been written on the relationship between people under 35 and their cars, but one thing I keep coming back to when I work with this group is the notion of a car has changed. The old images conjured up by Ned Jordan for the Jordan Playboy in the Saturday Evening Post no longer cause the imagination to fly.3 The car has become more of a high maintenance appliance. A very comfortable and useful one perhaps, but still an appliance that is not as central to your life as your smartphone. It should be an accessory to your phone.
A few years ago I visited a dealer to look at the current version of our then eight year old car. I've been happy with the car and its relatively clean interface and delightful shifter and clutch. The new model looked dramatically better on paper, but a test drive told me I couldn't live with the user interface. I made the decision then and there to do some work on our car with the intention of keeping it at least five or six more years.
I understand why the car companies aren't going in this direction, but they are opening themselves up. Someone, perhaps a Chinese, Indian or Korean company, could build a car that made sense as a smartphone accessory. While it may not be mechanically brilliant there may be a generation that considers that point irrelevant.
There are many other interesting forces in addition to this .. a lot of risk for some of them, but there are a few with little risk and nothing but upside.
So our cars are information producing objects and we're not terribly in the loop. There is much more information than we could use and the relevant question is to consider what we need and how it should be manipulated and used.
Recently a short piece by E.O. Wilson appeared noting that not all of science is intensely mathematical.
...
I was never more than a C student while catching up, but I was reassured by the discovery that superior mathematical ability is similar to fluency in foreign languages. I might have become fluent with more effort and sessions talking with the natives, but being swept up with field and laboratory research, I advanced only by a small amount.
Fortunately, exceptional mathematical fluency is required in only a few disciplines, such as particle physics, astrophysics and information theory. Far more important throughout the rest of science is the ability to form concepts, during which the researcher conjures images and processes by intuition.
Everyone sometimes daydreams like a scientist. Ramped up and disciplined, fantasies are the fountainhead of all creative thinking. Newton dreamed, Darwin dreamed, you dream. The images evoked are at first vague. They may shift in form and fade in and out. They grow a bit firmer when sketched as diagrams on pads of paper, and they take on life as real examples are sought and found.
...
While math is basic and essential for many areas of science, it is observation, logic, clear thinking and abstraction that is most important.4 In some areas there can be partnerships. Nothing comes directly from playing around with pure math and hoping it appears in the physical world. Nature doesn't make it that easy.
My background is very mathematical and it is a centrally important tool in my sport, but it is a tool. One learns how and where to use tools and importantly, how and where to use information.
There is a tendency to measure everything and try and use it - extreme quantification. The problem is the information and/or the models we have built are sometimes faulty. With careful thinking one can be find far more parsimonious.
In the mid 90s a few of us were involved trying to measure the human form to fit clothing. We came up with a scanner that could digitize the body with about a half million data points. A few years later I realized it was much easier to employ a good tailor who would take about 30 measurements. Furthermore, the experienced tailor, could communicate information to the pattern maker and cutters that was much richer than our big and somewhat noisy datafile.
It is possible to find examples of this everywhere. Currently the "big data" approach often misuses information. There is a notion that if you have enough "data" with the right sort of connections that you can mine that. To some extent you can, but it is not an easy task. There needs to be a deep understanding first and many lack that. I've seen a few brilliant solutions and many more expensive and troublesome toys that offer misleading information.
People have been doing smart and dumb things with data for hundreds of years. Computers allow us to do even smarter and more stupid things.
Imagination, day dreaming and play are central to getting the core understanding right.
___________
1 Some dealers update the flash memory of some your car's computers during regular service intervals, but this is rare and is often avoided even when the manufacturer recommends updates.
2 All American cars since 1996 have these ports - there is a lot of error code information as well as real time performance information.
3 Check out the second half of this post for details on the brilliant ad copy by Jordan.
4 A smart high school student can grock special relativity with no more than a bit of algebra and trigonometry and a bit of careful reasoning.
___________
Recipe Corner
A few weeks ago Om posted something on Holi which brought back a flood of memories when I lived with a group of Indian graduate students at Stony Brook. One of the foods I remember from Holi was a spiced milk - I think it was called Thandai.
I've had it with whole milk, but imagine it would be interesting with coconut milk and would probably also work, but be different and more healthy, with almond milk.
Thandai for Holi
Ingredients
° 3 cups whole milk
° 5 tbl white cane sugar
° 1/2 tbl fennel seeds
° 1/2 tsp black peppercorns
° 12 green cardamom pods
° 3/4 cup raw almonds, peeled and slivered
° an inch of cinnamon stick
° 2 tbl rosewater
° crushed pistachios
Technique
° roast the seeds, cinnamon and spices in a small pan without oil over a low heat until they become fragrant. Let them cool
° remove the cardamom seeds and discard the pods
° put the almonds, spices, seeds, sugar, cinnamon and rose water in a bowl and cover slightly with water. Let it sit for about 2 hours.
° pop it in a blender and grind until you get a paste
° add the past to the milk and mix
° filter through a cheesecloth into a bowl. (Rohini had a very fine mesh)
° chill it in the 'fridge, serve in chilled glasses and top with pistachios
Recipe corner
Steve:
My Ford uses my iPhone as the data connection layer. It connects to my phone, and has a siri-like interface. It downloads my address book, and I can ask the car to call people (on cell or home or work).
Additionally, it will play music (shuffle, play music by a certain artist) similar to Siri.
When I need directions, it will dial a call to a central service, ask me an address, download the directions to the car. If I deviate from the course it will ask if I want to update the route, and will reconnect and do so.
In that sense, it is using my phone in the simplest way, as a data connection. But I see greater potential in it...
Posted by: Howard Greenstein | 04/08/2013 at 07:14 PM
I have almost no use for a car living in København. People my age either have no car or something very small and rent for the three times a year when they need one. Cars are not sexy like they were for our parents. The idea of a sexy car is for old people. Maybe a motorcycle for some of the guys who love mechanical things.
Posted by: Jheri | 04/09/2013 at 09:37 AM