These are the extended notes from a talk I gave at #IoTlondon on January 2012. I have included some notes that didn’t make it to the final presentation due to time and format constraints.
The purpose of this talk was for me to introduce the motivations and initial prototype of Hintsights, a startup company I have recently setup and will be developing in the coming months.
I will start my presentation giving you a quick overview of my interests in The Internet of Things (IoT). Then I want to talk to you about the motivations and execution of Hintsights a new startup in the IoT space I’m setting up.
I’m a designer I’ve been professionally involved in interaction design for about 15 years. For the past 7 years I’ve run with my wife Artificial Tourism, a small independent design studio here in London.
These past few years we have worked on a number of projects you would identify as being part of an Internet of Things. Many of these projects have focused on the deployment of sensor networks in buildings for environmental and behaviour monitoring.
We are particularly interested in the development of tools to help people engage with information stemming from these networks, in buildings. We work with commercial clients , academics and multidisciplinary research projects.
I’m currently on the final stages of writing my PhD dissertation. My work is on Human Computer Interaction and the topic of my PhD is Persuasion and Artificial Intelligence. I design and build intelligent objects and spaces, and measure the effect they have on people’s behaviour.
This image is from a smart pop-up wine shop I built last year. The shop learned from its customer’s choices and recommends popular wine using light embedded in its furniture. People could also scan bottles to get 2-3 wine recommendations related to their choice, creating an implementation of a recommender system in physical space. Something similar to the ubiquitous “people who bought this also bought” statement that we find in online retailers like amazon.
Hintsights is a service and a range of products that revolve around the idea of a feedback network.
I think feedback is the “missing ingredient” or the “killer app” for the internet of things, and that is what I want to talk to you about today.
For most people IoT is still a fairly nebulous idea incorporating an amalgamation of technologies like cloud computing, sensors and actuators, automation, agency or adaptation, analytics etc…
However for most people (people outside the tech world) these tools have not real direct value or applicability. They are the building blocks of IoT, not the end product.
The end product is, I believe, more along the lines of products and services enhanced by connectivity. These would initially be not much different to many of the things we are already used to. However adding connectivity and feedback will make them part of an IoT and that will bring a radical change to our daily lives.
In order to understand how products and services enhanced by connectivity will change people’s behaviour and lives we need to understand how people’s behaviour changes (without them).
Here is a simple model that presents the elements you need to understand to change people’s behaviour: Choices, Decisions, Actions = Behaviour.
This can be best understood reading it backwards: Behaviour is an action driven by the decisions we make under the constraints of the choices available.
Examples include: beer at the bar, phones, cars, houses, careers, friends, etc, etc…
Much of the research and applications in this area concentrate in how choices can be shaped. For instance the idea of “choice architecture” or “nudges” in behavioural economics. Research in decision making in cognitive psychology focuses on different views of rationality, “biases” in our decision making and heuristics we use in specific circumstances. Designers shape choice in products and services by creating opportunities for interaction or affordances
Next, I think is interesting to look at how connectivity alone changes our behaviour. If we describe how the network (the internet) has changed to our lives using our behaviour model we can say:
Connectivity between people has increased our choices and brought huge flexibility in how we decide to connect with others. In other words we have the power to exercise choice and the flexibility to decide when, where and how to connect with friends, family, co-workers and customers. This has brought huge change to our lives that is reflected in the way we work, travel, shop, bank, etc…
IoT provides connectivity between objects, between people and objects, and people to other people through the use of objects – which I think is its most interesting proposition.
For instance IoT changes how we exercise through a running shoe. This technology enables us to get feedback on how we are doing, and the effect running has on us.
It can also connect us with other people that use the same running shoe. We can then get feedback from the network on how we are doing in comparison to others, our distance, time, bps etc… The same idea is currently being applied to other products like Wi-fi weighting scales, exercise monitors…
Is worth mentioning that I am not only referring to objects as immediately tangible or graspable products. After all a room is an object within a building, a building is an object within a street and that within a neighborhood, city… Equally a train is an object within a transport network. Learning how these “objects” are used brings huge potential for new services in the built environment powered by IoT and feedback.
Now that we have a basic model that helps us discuss elements in behaviour and we can see how the network (connectivity) changes them, we can start to think how products and services enhanced by connectivity will change our behaviour.
Looking at how IoT chnages the model above we can say connectivity enhances our choice resulting in improved decisions and a positive change in our behaviour. In other words, the key to IoT’s proposition is choice enhanced by feedback and how connectivity (between people through objects) creates new opportunities for interaction with objects in our daily lives.
IoT won’t increase our choice of running shoes or trains but it will give us a more informed, enhanced view on when and how to run or use the train which will improve our experience. That can only be realized through feedback.
If feedback is how IoT will enhance our choices and bring change to our lives is useful to see a cross selection of how we currently receive feedback from devices. It is interesting to note in the image above how presentation of data is mostly focused on dashboards with graphs delivered on computer or phone screens.
Graphs enable us to think retrospectively about a situation (our run, sleep, diet, energy use etc…) and analyse it in a way so it can be improved next time around. The downside of this form of feedback is that by the time we engage with information we are far removed from the point of choice or decision making, reducing our chances to act. Graphs are tools for data analysis, not for communication. We can all read a graph and spot the peaks and troughs, but that only enables us to formulate hypothesis or questions to be answered afterwards – unless the graph already depicts the output of analysis representing the answer to a question.
“The purpose of computing is insight, not numbers.” — Richard Hamming
I saw this quote on a tweet as I was editing this post and thought it would have made a good slide here.
I am not saying all retrospective analysis is flawed, but we need to be more selective in how we use feedback, we cannot engage analytically with every piece of information we encounter. Instead we need to find ways to enable people to benefit from data without becoming analysts.
The exceptions on the image above are Wattson & Joule, which complement numerical displays with ambient information. These devices provide multiple forms of feedback, cumulative views of data but also notifications pertinent to specific situations or events. If your energy consumption spikes, Wattson projects a colored light under it, signifying the change. The light provides an immediate, peripheral cue that enables you to change your behaviour in relation to your consumption, when and where it happens.
With the prospect of 50 billion devices enhanced with connectivity we need to start to think carefully about new ways to engage with them in a way we can reap the benefits of enhanced choice.
I believe there is a lot of room for innovation in this space.
That is where Hintsights comes in, our aim is to give you tools to benefit from data. Providing hints from insights, that make information easily actionable.
Using our service you can consolidate multiple information feeds in one place and get the intelligence you need.
Pip our information displays bring information to you when and where you need it. Giving you the gist, not more, so you can act or carry on with your life.
Our cloud based service allows you to monitor data feeds, and select when and how you want to know about your data.
Data feeds may include the amount of twitter mentions of your business, the pachube feed of your home energy consumption, your baby’s sleep patterns, local bus schedule, your work inbox etc… Anything that is web accessible in a standardized format, xml, json, csv.
You can then specify what are you interested in and when you want to be notified. Are your twitter mentions mostly relevant between 9am-5pm? Is your energy data most useful when you are at home in the evening? Maybe you just need to hear about your bus schedule, weather or traffic between 7am-8am before you leave home?
How do you want to be notified? Do you want to read that tweet that mentions you? A graph showing your runs for the week? Perhaps an icon alert when your Pachube energy consumption feed goes too high? Or a visual cue when you should go to sleep.
Your data will be delivered to your place, on your terms, with Pip.
Computers for places
This is Pip, I like to think of them as computers for places. You can place Pip on your living room, kitchen, hallway, office, or a coffee shop counter. Pip can connect to the network using a usb cable to your computer or wirelessly to your local wi-fi. Pip will connect to your hintsights account and bring seeds of information to you.
Pip has several display modes enabling to access data in the way that best suits you.
You can choose from a range of graphs, abstract images, an ever growing collection of icons or as scrolling text messages.
Where will the feedback network materialize first? There are a number of areas where we are building applications, these include Social Media, Environmental monitoring and Healthcare. See an added breakdown of scenarios of use below.
There are currently very few channels for places to engage with social media. Sure, people can checkin into foursquare and read reviews about a place in google or whatever. A live stream in place reflecting the impact and presence of that place in social media is a different story. Imagine you walk into a restaurant and it looks empty, you’re not sure, is an odd time of the day. Would you rather pull your phone, check google or foursquare for reviews to see what people think of it, or see a realtime stream from Facebook, twitter, foursquare etc… displayed with Pip and visible from the outside? Something like “zagat rated” but with live updating from social media.
How about feeding back into an office space the impact their work is having on twitter? Followers, mentions etc… Bringing social media buzz to the workplace. What if you had Pip plugged to Github? or displaying the status of your servers in the cloud? Got on Mashable? a surge of website hits? A virtual machine in your cluster going down? Be notified in realtime.
Utilities & Environment
Energy consumption awareness has become in recent years an important topic and a developing area of business interest. With several initiatives at national level to implement energy monitoring in households the market size and opportunity is ever increasing. Currently the offerings in this space are often in the form of tightly coupled products that enable you to sense and display data. Soon as we become concerned with the consumption of other resources like gas, water, air quality etc… will we still buy into a single closed platform per resource? Or are these likely to decouple sensing and displaying of data and converge into web services that will benefit from large distributed data analysis pipelines? How will we engage with that data? How about when smart grids become a reality in a not too distant future and we start producing as well as consuming energy? Will it not be useful to know when is more efficient for us to use and consume resources?
Heard about citizen science? How can you benefit from it? Have you looked into the size of Pachube’s community worldwide?
Healthcare & Wellbeing
Wellbeing has become one of the fastest growing areas for emerging technologies coming to market to make a real difference in people’s lives. The products mentioned above, nike+, fitbit etc… are the canonical examples of a “health2.0” revolution and a quantified-self movement worldwide. The promise? Proactive, preventive, healthcare instead of reactive palliative medicine. The platform? Feedback.
Thank you, @gonzillaaa @Hintsights
The event was a very interesting experience for me, I think Pachube is managing to create something that is quite hard to engineer intentionally. A growing group of people getting together informally and regularly to discuss projects and ideas in a open and welcoming environment.
On the day I was asked several questions and was approached by several people afterwards also via email. Here is a summary of the answers to the most recurrent questions:
What’s the plan? what do you have? and When will it launch?
I have a working version of the devices that I demo on the night and have demo elsewhere. Pip is an Arduino based platform using a RGB matrix display controller and a wireless transceiver put together.
I have written several libraries (c++) that include: a firmware for the device that incorporates a low level graphing API, a serial protocol to receive and plot data, display icons etc… I have a set of client libraries (python) that enables me to pull data from various sources (twitter, pachube and others) manipulate it and display it in realtime.
The service is currently sitting on a server on the linode cloud, and I am building it one application at a time (with Django).
I am collaborating with people and seeking partners to finalize the PCB and manufacture the electronics. Since I am working on my own, while writing my PhD some things move faster that others.
My timeline is:
- In three months approximately I’ll have a “developer friendly release” including firmware on github, client libraries to run locally on github and a pcb design ready for manufacture. I am looking at ways to either bring the project to kickstarter to finance the manufacture of a short run of Pip, or build a small online shop and allow people to pre-order and batch build the devices.
- After this point I will concentrate on developing the server further and having sold a few devices I will be in a position to proof market interest. This will enable me to raise finance and build a (non-developer) version of the service and devices to place in people’s homes and businesses.
Interested? Want to get involved? I am actively seeking partnerships. I am interested in data providers, businesses that believe their user base can benefit from accessing their data in a different way and want to explore ways in which we can collaborate. I am also seeking partnerships with hardware designers and manufacturers that are familiar with the Arduino ecosystem and can build devices for me. I will create the applications that will make customers come to you!
Developers? keep an eye on github or get in touch.
Customers? Want to buy one? Tell me what you want to use it for and if you’re happy to run code in the terminal and will strive to get one to you as soon as I can!