open search menuopen langage menutoggle menu
Actualités t3://page?uid=116

Frugal, IoT are words that go together well …

In this article by Gabriel Raymondjean and Gérard Barbosa, learn why taking a frugal approach makes it possible to get IoT projects out of the box and to achieve them successfully.

There is a strong wind blowing through IoT projects, in particular for the city, construction, energy and mobility sectors. The time of experimentation and other use cases seems finally to give way to "in the flesh" projects. A few examples:

  • increased mobility with better management of our urban mobile objects (scooters, bicycles, scooters and electric cars),
  • connected stations and airports to improve the fluidity of passenger flows moving around the station and improve passenger journeys even outside the station space (simplified access for Taxis, Uber, etc.),
  • energy management (electricity, water and gas) natively a pioneer in the field finds a second wind with optimisation offers aimed at individuals enabling all information from connected objects located in a home to be captured

A shudder was already perceptible at the end of 2019. Above all, the period of confinement, which we know has advanced attitudes on digital technology and teleworking, will probably have triggered the adoption of remote control systems. So much the better!

But let's be clear. While the rise of the IoT has been announced for over 6 years, concrete achievements to date remain modest (with the exception of connected objects and internal applications at industrial sites), this is not without reason. There are two main obstacles to mention.


The first obstacle for IoT projects is the weight of investments.

Indeed, an IoT project is first and foremost an expensive infrastructure and equipment installation project, with structural work, many intermediaries and logistics to coordinate for the supply, installation, then maintenance and evolution of sensors. The latter, moreover, as well as the digital platforms that supervise them, and the exploitation of data represent only a marginal share of the total cost of a project.

Worse, the value generated often struggles to convert into ringing and tripping currency. The first benefits are first expressed as well-being for the community, or result in a general optimisation of resources, which benefits everyone, but has no attractive return on investment. For example, the installation of new sensors in the candelabras in order to optimise energy consumption or the safety of our roads - but also in order to make the data collected available to local authorities without other services to individuals - clearly demonstrates the difficulty of generating wider uses that benefit installers.


The second obstacle is complexity.

First of all, we know that the multi-layer combination between construction, electronics, IT and networks, essential for any IoT project, is not simple. In this respect, the deployment of Linky (a Smart Metering project in France) is a success model, which must not, however, make us forget the difficulties of the early days to control the complete information chain.

The proliferation of technological options is a second complexity factor, which plays on the choice of the type of information sensor, the consideration of obsolescence risks, and the choice of suppliers, all amplified by a lack of standardisation.

Three other complexity factors, far from being the least, are still to be taken into account:

  • Coordination of public and private actors, involved in projects lasting several years.
  • The security challenge arises at all levels of the information capture and processing chain.
  • The necessary change support during IoT deployments in spaces connected with the public, given any current societal reluctance.

Of course, faced with these obstacles, efficient governance, informed technical choices and solid control of project management are essential. But in our view, it is by taking a frugal approach that it becomes possible to get projects out of the box and to achieve them successfully. Put simply, a frugal approach is to find creative solutions to "do more with less", and often to make possible projects that, handled in the ordinary way, would be ruled out because they are too expensive. A frugal approach is based on the box: value, costs, creativity, iterations.

In the case of IoT, frugality can be broken down into at least 5 areas.

  1. Respond to the real need that generates value, or else give up
  2. Achieve a result as quickly as possible, because there is an induced benefit
  3. Draw on the existing situation and accept compromises
  4. Optimise the acquisition-processing of information pair
  5. Guide technical choices based on value rather than performance


1- Cardinal point of any frugal approach, the deconstruction of the need, aims to clarify the nature of the expected value and the levers that will make it possible to obtain it.

This compass is essential for the IoT because nothing is really as simple as it seems, and the multiplicity of possible responses quickly led to diverting from initial objectives and complicating things. The use of cameras, escalator sensors, or lifts, mobile phone tracking are all possible solutions for dealing with taxi queue management and the management of flows of people in and around stations.


2- The ability to iterate quickly and improve the response incrementally is also an essential point.

If, for example, to optimise flows of people, it is necessary to count the number of people present in a room, then it may be wise to first set up connected carpets instead of using video data. Their very rapid implementation will provide easy-to-use information that will allow the flow management algorithms to be tested in the moment. The processing of video data, which is more complex to implement and subject to regulatory constraints, may take place subsequently, only if necessary.

Here again, the obstacles to obtaining a satisfactory response, because many and unpredictable, favour a trial/failure approach, and a moderation of the expenses incurred at each stage.


3- Exploit the existing situation and accept the compromise.

Even if we are only in the early stages of the IoT, the ground is already occupied. Many sensors and networks are already in place everywhere, whether they are cameras, thermal probes in air treatment networks, position or speed sensors on lifts, etc. all the data available, reliable and often free of charge. So that before any design phase, it is a good idea to look at what already exists and be creative to make the best use of it… Having deconstructed the need, as indicated above, will be a valuable help in validating creative solutions.


4- Choose your priorities between sensors and processing.

If the IoT is an acquisition-processing-decision chain, its value will be constrained by the smallest of its components. While each link benefits from uninterrupted technological progress, not all move at the same speed and the costs of change are not the same. The search for the best compromise must be made by considering both the short and medium term, in an open mind with regard to future innovations but which are still unknown. It seems reasonable to invest preferably in the means of information processing, as they can compensate for the shortcomings of average quality data, and prove less costly than sensors. However, there is no absolute rule, and some progress "by qualitative leap" may warrant an overhaul of the entire chain.


5- Still in the cost register, several technical parameters have a significant influence, both on investment costs and on operating costs.

The amount of data to begin with. This is an important lever both for estimating the power (and therefore the costs) of sensor processing and decision-making as well as the necessary network and storage costs. As a general rule, we often have too much data that we do not use enough.

The choice of standardisation (processing platforms, communication protocols, etc.) rather than specific solutions has a priori favourable impact on costs and works in favour of the iteration speed mentioned above.

Finally, energy consumption must also be taken into account, even if its effect is most often at a lower order of magnitude compared to the other parameters.


While the crisis that is beginning encourages us to reduce costs, taking these dimensions into account in the context of a frugal approach must enable companies and local authorities to find a way of achieving their projects, which delivers the expected value, at a controlled cost and which lays the foundations for the digital transformation and the ecological transition.