The HUMANE method to support the design of human-machine networks

In the networked society, the people and the machines we rely on are approaching a pattern of always on, always connected. Our activities at work and in our private lives increasingly depend on the networks of humans and machines. The proliferation of technology and increasingly complex interplay between humans and machines bears significant challenges for the design of what we refer to as human-machine networks (HMNs). We expose and address the challenges via the HUMANE method.

We observe that the successful outcome of highly diverse activities such as industrial innovation processes, sharing economy transactions, citizen science, and news verification practices are increasingly dependent on HMNs. However, for European industry and public sector to benefit from efficacious HMNs, we need to understand the key characteristics of such networks, and how these characteristics in turn affects people’s experience and behaviour in the networks, and how the networks evolve. For example:

  • How does the increasing reliance on machine intelligence in HMNs affect privacy and trust perceptions?
  • How does the centralized organization of platforms for open innovation or citizen science affect motivation to contribute?
  • How to ensure constructive collaboration between humans and bots in platforms such as Wikipedia?

A key assumption in HUMANE is that the purposeful design for HMNs will benefit from understanding the characteristics of the HMN in which it is to support. By complementing socio technical systems design and human-centred design, we seek to understand HMNs not only as particular and contextually embedded, but as holding a set of generic characteristics which may apply to HMNs across application areas and domains. We have captured key characteristics within the HUMANE typology, which includes the network actors (human and machine) and their relations, as well as the network structure and extent.

We also need new approaches to the design of HMNs. Through traditions such as socio-technical systems design and human-centred design we are provided frameworks for analysing the context of technology design and processes for involving users and stakeholders in the design process to ensure effective and efficient work support. However, designing for networks of humans and machines remains challenging.

Examples studied by HUMANE researchers include online innovation platforms that fail to strengthen innovation capabilities (Lüders, 2016), how procedures for including and excluding people in online collaboration may work against their intentions (Rudas et al., 2017), and how the excessive use of bots in a network for collaboration may change or, at worst, challenge collaborative culture (Tsvetkova et al., 2017). Failing to design for purposeful participation and interaction within HMNs may threaten both the value of investment in a design, but also potentially represent lost opportunities to improve on the quality and competitiveness of European society.

To make the HUMANE typology actionable for analysis and design for HMNs, we have developed a five-step process for characterisation and analysis of the HMN, as well as the transfer of design knowledge and experience. The process is referred to as the HUMANE method, which includes the following:

  1. Context and scoping: Explicate the purpose and objectives of the HMN.
  2. Network characterisation: Establish the HMN profile through the typology dimensions and identify similar HMNs.
  3. Implication analysis: Analyse key implications of the HMN profile in terms of user experience and motivation, behaviour and collaboration, innovation and improvement and/or privacy and trust.
  4. Design considerations: Cross-domain identification and utilization of design knowledge and experience through a design pattern approach. Design knowledge is shared through what we refer to as design considerations.
  5. Evaluation: Assess and refine current design proposals on the basis of the input from the HUMANE approach.
humane-method-hcd-mapping

Overview of the HUMANE method, mapped to the phases of the human-centred design process.

The typology and method is intended as a supplement to the human-centred design process (ISO 2010), addressing analysis, requirements, design, and evaluation. Specifically, the typology and method is intended to support initial strategic reflections and considerations at the outset of the design process. The typology and method have been applied and evaluated in eight collaborating cases within the HUMANE project. However, this is not limited to the design process. One of the key benefits include facilitating cross-disciplinary communication, providing a framework for discussing and understanding HMNs in general. For example, it has been used as part of a process that has led to three policy-focused roadmaps for future HMNs in the areas of the sharing economy, e-health and citizen participation.

For a more detailed overview of the HUMANE typology and method, including an application example, please see our method white paper.

Engen, V. and Følstad, A. (eds) (2017). The HUMANE typology and method – supporting the analysis and design of human-machine networks. Zenodo.

References

ISO. (2010). Ergonomics of human–system interaction — Part 210: Human-centred design for interactive systems. Geneva, Switzerland: International Organization for Standardization.

Lüders, M. (2016). Innovating with users online? How network-characteristics affect collaboration for innovation. Journal of Media Innovations, 3(1), 4–22.

Rudas, C., Surányi, O., Yasseri, T., & Török, J. (2017). Understanding and coping with extremism in an online collaborative environment: A data-driven modeling. PLoS ONE 12(3): e0173561. https://doi.org/10.1371/journal.pone.0173561

Tsvetkova, M., García-Gavilanes, R., Floridi, L. and Yasseri, T. (2017). Even good bots fight: the case of Wikipedia. PLoS ONE 12(2): e0171774. https://doi.org/10.1371/journal.pone.0171774