Robotics: the great confusion

The proliferation of software packages, fads and the absence of standards are creating a joyful confusion about digital in general and about tools such as RPA (Robotic Process Automation) or Artificial Intelligence in particular. It is not always easy to apprehend the service really covered by these new softwares: information gathering, internal or external integration (Blockchain) of processes, process piloting.

 

Robotics, IT Resilience and Operational Efficiency: three closely interrelated topics. With competition from GAFAM and other Fintechs and pressure on revenues, cost control must remain a priority. The Financial Majors continue to benefit from the size of their customer base, product portfolio, resources and skills, and they know the meaning of the word "compliance", but this business is exposed to slow erosion. The RPA can be a strong incubator for productivity and innovation: Managers can only be seduced by these new automatons presented as inexpensive, easy to implement and allowing rapid savings. These 'new' technologies are promising, but they must be used wisely, the ground must be prepared and the architectures must not be mortgaged.

 

These robots can replace humans and quickly fill integration gaps by avoiding recopying, with possible automatic complementation with data retrieved from other databases or even on the WEB. "Gentil Robot" has a strong capacity for data analysis, enrichment and classification (rule engine, parameterization at the user's fingertips via metalanguage, document archiving) but, as you will have understood, process integration can be an alternative!

It is difficult for a robot to work if it does not receive structured food, but some are better than others on this terrain and have natural reading or voice comprehension skills. Artificial intelligence implies an ability to learn, visionaries speak of 'artificial curiosity'.

These new technologies are a priori easy to deploy, due to their strong adaptability to a heterogeneous context, be it operational, organizational or technical: this quality could be a poisoned gift if RPA remained only a comfort and symptomatic drug, masking serious pathologies that will continue to metastasize in silence. Demonstration.

 

 

Many companies remain trapped by obsolete solutions that are difficult to get rid of because of a loss of control and a lack of visibility on dependencies and integration constraints upstream or downstream. Resilience is difficult to sell and program into an IT master plan because of the lack of directly measurable ROI.

A heterogeneous and ageing architecture induces a profusion of skills, financial costs (licences) and structure: the RPA could be just a smoke screen, adding an additional enclosure to this technological prison. The implementation of a robot cannot be done without in-depth work on the data.

 

 

The digital revolution and the economic climate require a rethinking of organizational models to make them more flexible and responsive. Hierarchical and matrix models, with their inertia and silos, are being replaced by "neural" models, with variable geometry, federating teams according to a logic of projects managed in agile mode. 

 

 

Robotization will not magically solve the structural weaknesses of your processes, i.e., defects that could degrade their efficiency, production deadlines or the quality of the results. Automating sick processes may only result in an even faster degradation of quality. Vigilance especially on long chain, cross-departmental processes.

Data quality and value chain cleansing are prerequisites for automation. Process optimization can take place in many dimensions: standardization, fluidity, security, shortening, integration, defragmentation, quality and availability of inputs, etc. It is necessary to avoid that RPA perpetuates bad habits or activities that do not bring any added value.

 

 

The best candidates for RPPs are processes that handle high volumes of transactions, following a pathway of and a stable and standard operating procedure, applying precise rules. In addition to the continuous improvement of the organisation and processes, the action plan should be based on three main lines of action:

1. A pilot project. Of course you have to surf on this "4".^th A proactive approach is required in order to take advantage of the "first-mover advantage" when launching new products. It is not possible to wait for the clean-up of all critical processes: the "learning by doing" remains an option, and if the pilot process is well chosen and  if the target operational model is well defined and accepted, a flagship, tactical, "agile" project can serve as a "proof of concept" and catalyst.

 

2. A master plan. The digital transformation must be accompanied by an in-depth evaluation of the processes to map them, assess their maturity and launch, if necessary, re-engineering projects, by equipping oneself with the methods (Lean Six Sigma?) and skills required. The identification and measurement of manual coding activities are obviously prerequisites to any approach.

It will probably be difficult to select the Swiss Army knife that will solve all integration problems. There should be no mistaking the interpretation of the needs: automation, process integration (avoid copying), just-in-time with the Market (blockchain, bis) data conversion (structuring, codes), new customer services, new sources of revenue, cost reduction. You need a master plan to identify targets and set priorities.

 

3. Managing resistance to change. Success will also be achieved if the approach is accompanied by a communication plan highlighting the strategic priorities and the deployment plan and aimed at mobilising energies and imaginations on the choice of the tool and the best way to use it. The appropriation of solutions is a vital factor and the risks of active or passive resistance to these disruptive transformations should not be underestimated.

 

Bernard Timmermans, UP' Magazine Guest Columnist, Independent Consultant