Not 3D printing houses after disasters

I get emails all the time from people hoping to solve the post-disaster shelter and housing problem by 3D printing houses.

Instead, if you have to 3D print something, how about:

  • 3D printing cheap probes to test cement block strength (like this one)
  • 3D printing demonstration model walls/miniature rebar systems (lego-styli)
  • 3D printing long lintels with neat interlocks into block or adobe
  • 3D printing bespoke gusset plates and straps for any old type of homemade truss


Enabling rapid post-disaster repair: innovative institutional decisions

UCL and Habitat for Humanity were unsuccessful in their invited resubmission to the Humanitarian Innovation Fund Large Grant Facility.

Here is the proposal for anyone who wants to take it forward as well as the feedback from the HIF panel.

"Your application was well scored by most in the group of reviewers, however there was one member who did not consider the project wholly innovative, this reduced the overall scoring and therefore was not shortlisted. Hopefully the comments from the reviewers will be of use to you:

  • ‘Although the promotion of post disaster housing repair and reconstruction is not in itself innovative, an approach to changing knowledge, attitudes and practice in an integrated way for relevant for ALL actors is a new area of research.’
  • Relevance / Impact: This project is highly relevant, given criticisms of current shelter approaches. This seeks to realistically address issues of focusing more on (developmental approaches) of repair
  • Methodology / Approach: the timeframe of 12 months is short, 18 months would be better
  • Feasibility: Clarity of intent is there
  • Team Composition / Capacity to implement: A good team that combines practice with academic rigour

The proposal has been thoughtfully revised, and framing the project as assessing the need to change knowledge, attitudes and practice is now a major strength. Two major concerns remain regarding the proposed activities.

  1. The focus on Haiti for the case study. Although the post 2010 earthquake shelter response in Haiti was fragmented, this was exacerbated by the exceptional scale of this disaster, lack of a functioning government and a regulatory environment, the significant resources available, and the multiplicity of operational agencies. Within a year a more coherent approach to housing recovery had been widely adopted, including repair, rental etc. This project would benefit from identifying one or two more typical medium scale post disaster contexts, with a functioning Govt and regulatory environment and modest recovery resources.
  2. The project also proposes addressing the institutional decision-making challenges within Habitat for Humanity. As HFH are a housing organisation, focussing on housing development/construction, they are an atypical representation of the wider housing sector. For the project to have sector wide benefit it would be beneficial for the project to address the internal decision-making processes with agencies for whom shelter/reconstruction is but one of the sectors they focus on, and not necessarily a core competency."

HIF, February 2014


Augmented reality apps for assessment and/or diagnosis of building damage after disaster 

Key challenges in evaluating large scale damage to buildings after earthquakes are:

  • observing damage because street surveys cannot take into account internal damage;
  • estimating damage because damage can be overlooked or underestimated, depending on the quality and orientation of the photographs, especially for the lower damage grades;
  • scale because traditional assessments depend on expert analysis and the number of people on the ground with the right experience and training may be limited;
  • doing assessment remotely depends on reliable data (ie good photgraphs) and
  • these approaches often miss the opportunity for helping "non-expert" citizens to learn about damage types and diagnosis as they take photographs of their homes and neighbourhoods.

Three cool things to do about this would be to:

  • develop an app to help people take "good photographs (of damage)" and
  • build into those apps any relevant learning/information on basic mitigation and principles for building back safer
  • build on platforms that do this already (i.e. AirBnB's photo repositories and their network of local photographers...)

Not designing an innovative transitional shelter

I also get lots of emails about how to design the most efficient, cheap, lightweight, quick, easy to use post-disaster shelter. In many situations, this is probably not the technical problem that most needs attention: instead it would be much more interesting to have engineers think about the complex and risky decisions concerning:

  • the housing stock that remains,
  • supporting people to rent (AirBnB and the Government of Japan have been clever about this),
  • repairing housing,
  • shoring up slopes and restoring communal infrastructure.

If you still think transitional shelter gets a lot of airtime and must therefore be a technical problem worth solving. Think about this...

The term was intended to describe the transition process that people go through after a disaster: if you’ve lost your home and you flee to another area or you are displaced or you seek temporary evacuation in a centre or you stay with a host family, you’re in a state of transition between finding some suitable alternative accommodation and being able to return and rebuild or repair or resettle somewhere where it’s going to be easier for you.

So, transition is a process (rather than a defined period of time) and is about the transition of different groups of people (not just the transition or upgrading of a shelter). This helps to frame the post-disaster shelter challenges for a government as strategic questions rather than purely technical problems because you have to start looking at:

  • the different situations that different population groups find themselves in,
  • how many people are in each of those different situations,
  • who might be particularly vulnerable in those groups and then
  • what your menu of options might be - high level, flexible and equitable - to meet the needs of those people. a broad pallet of things you can do (or support) at a high level and how those play out at a slightly more detailed level of granularity.

Transitional shelter in disaster response is contested partly because there is a lot of confusion around the term. A key confusion (and barrier to making useful comparisons) is thinking of transitional shelter as only the pre-fabricated housing or the kit, this is just one component and disaster responses where pre-fabricated housing has dominated have been criticised because:

  • they’re limited in scale because they are expensive and you can’t reach that many people,
  • they are limited in scope because only people with a place to put such a thing can receive one,
  • they don’t influence the underlying culture of building and safer building
  • they don’t build capacity for people to recover for themselves because they are imported,
  • they don’t necessarily stimulate local economic recovery

This is the list that needs attention. IMHO.

Can we be more than the sum of our geeks

Work we did thanks to the Humanitarian Innovation Fund showed that most of the technical evidence gathered by engineers and NGOs after disasters does not add up to answer some of the fundamental questions about shelter and housing processes: who builds, when and how; why people live and build in the places they do; how housing is financed normally; what common defects might make buildings fragile; and what really changes building practices given that this depends on building culture, markets, training and skills, taxation, styles/aspirations, livelihoods, location and regulation.

  • What sort of information would support local decision-makers and households to accelerate their recovery?
  • What sort of information would stimulate the funding of strategies that fit the context and respond to prior housing processes?