Smarter solutions for a better coffee making experience (Part 2)

Let’s continue our exploration of the contextual intelligence domain and see how simple things in our everyday life could potentially be affected by this development. Our goal in this blog series is to design a context-sensitive, smart Nespresso Machine that can provide us with exactly the right cup of coffee at the right time and that can support us in performing simple maintenance activities.
In my previous post we looked at the variables of preparing a cup of coffee with a Nespresso Machine:

  1. Brewing ratio: ristretto, espresso or lungo
  2. Additions: sugar / sweets, coffee milk / coffee milk powder, steamed / frothed milk
  3. Flavor
  4. Intensity

Other important aspects like temperature and pressure were left out of the list, because they cannot be influenced with a setting on the machine.

Relevant Context Variables

So, what are the relevant context variables that influence our choice of coffee. In my opinion all of these variables can influence our choice:

  1. Taste
  2. Flavor / smell
  3. Intensity
  4. After-taste
  5. Caffeine
  6. Day of the week
  7. Time of day
  8. Location


Taste

In order to determine the taste preferences of a user we could map specific coffee tastes to more generic tastes and ask the user one time – during registration – for its preferences:

  • Spicy
  • Cocoa
  • Roasted cocoa
  • Dark Chocolate
  • Bitter
  • Fruity
  • Caramel
  • Cereal
  • Malted cereal
  • Citrus
  • Winey
  • Sweet
  • Flowery
  • Mint
  • Hazelnut
  • Vanilla 
  • Milk


Of course combinations of tastes are possible, creating an unique taste pattern for each user.

Flavor / smell

Like with taste, the flavor / smell preferences could be asked to the user during registration or initial set up:

  • Intense
  • Roasted cocoa
  • Roasted cereal
  • Woody
  • Flowery
  • Caramel
  • Citrus
  • Malted cereal
  • Mint
  • Hazelnut
  • Vanilla 
  • Winey
  • Sweet
  • Neutral

As you can see, some of the flavors / smells overlap with the tastes. Of course this is not an exhaustive list, it is merely a first attempt to capture the flavor / smell characteristics of different types of coffee.

Intensity

The intensity could be classified into five options:

  1. Very intense (viscous, dark black)
  2. Intense (dark black)
  3. Neutral (black)
  4. Mild (dark brown)
  5. Very mild (brown)


Caffeine

For the caffeine we can simply identify two choices:

  1. With caffeine
  2. Caffeine free


After-taste

Some people like to still taste the coffee minutes after they finish it, others want to avoid such an after-taste. For this example we can use a generic classification:

  1. Strong after-taste
  2. Mild after-taste
  3. No after-taste


Day of the Week and Time of Day

Some people have one coffee recipe that fits their preferences every day of the week. For others the weekend is the time that they drink Latte Machiato’s, while on working days they stick to regular coffee with added milk and sugar. The same is true for the time of day. Perhaps someone likes to drink coffee with caffeine during the day and caffeine free coffee after 5 pm.
In order for our smart coffee machine to be able to predict the right cup of coffee, it has to learn the habits of its users, see patterns and predict what is valid based on the available information.

Location

Another context variable that could influence our choice for a cup of coffee is the location. Although a coffee machine typically has a fixed location, I can see the benefit of a smart coffee app on my smart phone that can operate coffee machines at home, in the office, at my clients office, perhaps even in my car.

Combining all of these context variables into one smart solution

It takes the experience and knowledge of a barista to mix all of these variables and define the best coffee experience for each user and context. Combine this with a self-learning system that lets users evaluate their coffee after drinking it and improve its advices based on that to create a system that can predict for every user within every context how its cup of coffee should be made.
In this blog series I used the example of a smart coffee solution to illustrate the possibilities of Contextual Intelligence.  With this example in mind it is pretty easy to imagine how these principles can be used in the business domain to automate business processes based on context variables. To be continued….

Comments

Post a Comment

Popular posts from this blog

Using the Golden Circle Model in your resume and LinkedIn profile

Image
The last three years I worked as a (senior) business consultant and information architect at software company Be Informed in Apeldoorn, the Netherlands. We worked on an amazing product, a ground breaking model-driven business process platform. And my job was to explain the possibilities of this platform to various audiences. The Be Free program, which I launched in September 2013 was a marketing campaign that targeted on making a wider group of people familiar with the product. The campaign was a great success and I had the feeling that I was really helping people in proving their ideas to their customers and managers. But the interaction was on-line and from a distance, never close enough to really understand their business goals and issues.  The sudden end of my contract with Be Informed, forced me to rethink my professional goals and ambitions. What do I really want? Do I want to be a product marketer, focused on 'pushing' a product? Or do I want to be a business cons...
Read more

Improve Product and Brand engagement through smart documentation - A White Paper about the challenge of creating user-centric documentation

Image
For most product manufacturers, delivering adequate documentation is a constant challenge. The increasing complexity of products and the stricter regulations for product documentation often lead to extensive documentation sets with 500+ pages manuals on-line or in print. The question is: how effective is all this documentation? Smart, user-centric documentation Users can be overwhelmed by long, jargon-loaded user manuals, and frustrated by a lack of clear answers. When customers don’t find a solution, blame inevitably falls on the product and the company. Bad documentation - incomplete, unclear, out of date - creates a negative engagement with products and brands. Figure 1. Smarter documentation asks users what they need to know Smart documentation leads users to success and satisfaction. These are the users who share their positive experience with family and friends, and convince them to become customers in turn. Using semantic technology, we can create the smart document...
Read more

Writing effective articles

Image
I am a writer. I have been all my life and I wouldn’t know what to do if I would ever lose this talent. I have been writing about many topics: from welding procedures and high tech greenhouse technology to  software implementations and the use of semantic technology.  Apart from the writing I enjoy digging into these topics and finding out how things really work and what the benefits are of these new technologies or approaches. But is a talent for writing in combination with an interesting topic enough to form the basis for a successful and effective article? Or do we need more? In this series of blog posts I will explore the criteria for a successful article, not necessarily providing you with a fix set of rules but rather focusing on best practices, lessons learned and examples from my own work. On the cover of a magazine One example from my own work. In 2008 I wrote an article called “The Flexible Intranet” about using semantic technology to create smarter Intranet sites...
Read more