Section: Scientific Foundations

User needs analysis method for smart home

Summary

Ubiquitous computing promises unprecedented user services from the flexible and robust combination of software services with the physical world. Our hypothesis is that end users are willing to shape their own interactive spaces by coupling smart artifacts, building imaginative new functionalities that were not anticipated by system designers. Our work is concerned with the fundamental meaning (and human needs) of building confederation of interoperating smart artifacts. The Social sciences offer tools and methods for exploring human needs and behavior. However, the novelty of our problem requires to solicit participants imagination while at the same time controlling the experimentation and respecting the privacy of their intimate home. We set up the DisQo method [29] to explore how far people are ready to envision the interconnection of everyday devices to improve their lives. Results show that services suggested by our family members fall into four categories: Service substitution, Service improvement, Service chaining and Service "starter".

Detailed Description

Drawing on Davidoff’s et al. method and conclusions (i.e. "families want more control of their lives" [33] ), we focused on "busy" families. The participants have been solicited through bulletin board advertisements, email, as well as from personal relationships.

We are interested in determining how far people are ready to envision the interconnection of everyday devices to improve their lives. For so doing, we have used a combination of interview (good for clarification), playful cultural probe (appropriate for respecting privacy and for improving subjects involvement). The presence of the experimental team (ourselves, from 1 to 3 persons) was limited to 1h30 per family home. Fieldwork was structured as a four-step process: photographing, interview, game, and debriefing.

• Step 1:

  Photographing. Two volunteer family members were asked to take pictures of 10 objects at the rate of 2 objects per room. For each of the 5 rooms of their choice, they were asked to take a picture of one object that they considered to be necessary in their everyday life or that would help them in organizing their lives, as well as a picture of one object that they considered to be superfluous but valuable (typically, a painting). The volunteers (in general, the parents) were not supposed to be in the same room at the same time so that they would not know which pictures the other member had taken. Meanwhile, the experimental team would wait sitting at a place indicated by the parents (typically, the living room where they usually meet with friends and visitors).

• Step 2:

  Interview. We then conducted an interview with all the family members, using the pictures as input material. Questions were directed at understanding the reasons for their choices, the value attached to the objects or the services provided in daily use. Special attention was given to the (many) remote controller(s) typically found in the household environment. We progressively oriented our questions towards novel uses of smart artifacts. In particular, we asked which objects of the house (including those on the pictures) they would qualify as "programmable" (e.g., TV’s, washing machines, alarm clocks), "communicating" (e.g., computers, mobile phones), or emotional (i.e. carrying intimate value). This was used as a means to elicit routines and exceptional needs as well as to prepare the game developed in Step 3.

• Step 3:

  Association game. The association game drew on people creativity using the pictures as play cards. Pictures were sorted randomly and presented two at a time (then, three at a time) on the tablet PC. Family members were asked to imagine which service(s) and value(s) these two (or three) objects coupled together would provide them with. Random coupling was designed to solicit imagination in unexpected ways as solutions creativity grows with the semantic distance between elements [48] .

• Step 4:

  Debriefing and informal discussion. The last stage was dedicated to debriefing, including opened friendly discussions.

Overall, we have collected comments and objective data for 349 couplings for a total duration of 25 hours of our presence in the 17 family homes. We found a number of facts that are quite consistent with the results reported in prior literature:

1. “Wake-up” time, “on-the-way-to-home” and “arrivinghome” times are key moments to people. To save time and improve efficiency, activities are organized into wellpolished procedures. As a result, exceptions to these routine tasks are sources of stress. Support for avoiding or for solving exceptions is one class of services expected from a smart home.

2. With regard to programming, attitudes range from "I do not want to be assisted" to “It will work 99% of the time, but it will be hell for the other 1%”. Motivation for programming is systematically grounded on a clear straight forward observable benefit

Our data from the association game shows two important results:

1. Family members are prone to envision new services when coupling involves one “communicating” object, or one "programmable" object, at least.

2. The “communicating” capability has more impact than “programmability” on the capacity of family members to imagine new services.

The services suggested by our family members fall into four categories. We illustrate them with the most typical examples drawn from our fieldwork :

1. Service substitution. People have for instance observed that, for the same (sport) events, commentaries on radio broadcasts are richer than those provided by TV.

2. Service improvement. Some household appliances such as washing machines and storage areas, do not provide any convenient way to control and monitor their current internal state. Appliances than are not sufficiently equipped could be improved by coupling them with additional input and output facilities such as those of the TV set.

3. Service chaining. Service chaining is intended to improve comfort, wellbeing as well as resources for the routine, but hectic, activities. For example, ?picking up the towel after the shower would trigger the coffee machine so that coffee would be ready just in time, at the right temperature, along with the radio turned on in the kitchen broadcasting the news using the appropriate sound level?.

4. Service "starter". We have observed that some appliances serve as triggers for services that are expected to be precomposed to support routine activities. Not surprisingly, people also want to have an explicit and reliable control over the home (cf. the worry that 1time, the house would turn into hell).

Based on this results, we currently explores interactive systems that enable end users to programm their smart environment.