Is this your challenge?

Knowledge is an asset. Within a company, this might actually be the most important asset. For instance, the loss of a critical piece of knowledge can have a drastic impact on an organisation, either in terms of finance or productivity.

In a less drastic situation, identifying the right information in a large organisation is extremely time consuming. According to a McKinsey report, workers spend 1.8 hours every day looking for information. On a full year, it equals to more than 400 hours, i.e., 50 days …

In the nuclear industry, we are all too much aware of the importance of Knowledge Management. Organisations have many databases ranging from internal wikis to SharePoints, siloed and not talking to each other. Those systems are obviously useful and relevant but finding the right, non-siloed information takes time.

Although you have those systems at hand, most of the time the right information is known by someone within your organisation, and are not necessarily recorded. Finding that kind of information to answer your question is even more time consuming. So, how do you find the right expert? How do you get in touch with him/ her? Do you even speak the same language? All in all, you know that getting that expert and required information will take hours.

With the development of Artificial Intelligence (AI) and Natural Language Processing (NPL), there are ways to drastically help workers within large organisations to find the right expert with the information you need, quickly, efficiently across the whole organisation.   

inTechBrew’s insight

Incorporated in 2018, Ask for the moon, a French startup started to focus its value proposition on the shocking fact that 80% of knowledge is actually tacit.

The approach undertaken by this company is to both focus on Industry 4.0 tools such as AI and NPL whilst leveraging people who own tacit knowledge.

This combination of tools and access to experts is focussed on answering three great challenges when it comes to knowledge management:

• The technical systems designed today will still be operating when the original designers will be dead;
• The turnover within a company is now much higher than before;
• Most of the knowledge is actually tacit.

The approach Ask for the moon took is not to be another tool but rather being integrated in existing tools as an extension. The solution can be used in Teams, Chrome, etc., the daily software used by workers. It drastically increases the adoption rate.

What we like at inTechBrew is the focus on enhanced social network through AI and NPL to retrieve implicit knowledge. We are constantly looking for information. Although there are many existing solutions, most of the time they are siloed and not easy to access. With Ask for the moon, there is a real opportunity to find the right information, quickly and from the right expert. Ask for the moon unlocks this untold knowledge, not stored in any database which essential in an organisation.

What’s even more interesting is once the user is connected to the right expert, the information given by the expert is then stored in a database. Next time someone has a similar question, Ask for the moon will first look into the database before reaching out the expert, reducing the amount of solicitations an expert received.

A mix of social network data miner and database in a nutshell.

User case 1

World – On-going

With few thousand employees spread across the world, Framatome is fully conscious of its need to optimise its Knowledge Management.

Although within Framatome, there are many Knowledge Management databases, established processes and Good Practices, they are fully aware that a significant chunk of the knowledge has not been captured.

Their objective is to be able to retain information for at least 120 years (e.g., today’s knowledge must be available in 2140 when the Flamanville EPR will be dismantled).

To that extent, Ask for the moon is used as a Q&A platform for technical queries, connecting multiple experts across the world to capitalise on the knowledge.

The process within Ask for the moon was straightforward:

• Someone ask a question – which is characterised with an NPL algorithm;
• If the question is not within Ask for the moon database, the AI identify the right expert and connect him/ her with the person looking for answer;
• The expert provides an answer;
• The answer is used to populate the database.

Within Framatome, Ask for the moon confirmed the strong upside of deisolating knowledge transfer within the organisation from a geographical standpoint but also in terms of competencies – e.g. commercial team gathering information from technical experts to answer a bid.

The system was deployed within a month across the organisation and strong benefits were seen within a period over three months:

• 900 users (technician, engineers and researchers) during the first month all the way to 1,400 users during the last months;
• 907 questions were asked;
• The answers provided were seen 30,915 times;
• 85% of the answers are answered within a day;
• 2,478 hours of looking for an answer to a question where saved;
• Experts received 50% less emails coming from colleagues looking for an answer to a question;
• 95% of the users agree that Ask for the moon save them time;
• 91% declare that the app is not redundant to other existing solutions they have within the organisation.

User case 2

France – 2022

Cegelec CEM ran Proof Of Values to identify the value of Ask for the moon across the whole organisation. The trials took place over a three-month period. The results shown below are monthly averages.

With the AVENIR project – an internal quality improvement cross-company project -, Vinci Energies – the parent organisation of Cegelec CEM – has set itself the strategic objective of sharing and capitalising on knowledge. Cegelec CEM therefore has a strategic objective of sustainable development of knowledge by allowing an increase in the skills of its employees.

Preparing a commercial proposal for a technical tender in the nuclear sector

The commercial team and technical experts write together their technical capabilities based on past projects and documents.

For example, in one tender, 4 people were involved, 3 questions were asked and got 6 answers. 4.6 hours of looking for information were saved over a month.

For full-scale deployment, it would involve 108 employees with a targeted time savedof 52 man.days per year – not correlated with the POV values.

Process optimisation

The internal QA/QC team and the engineering office are writing internal processes and procedures. They identify the gaps and clarify the associated documents based on questions they get from the workforce.

On a specific process, 3 people were involved, 6 questions were asked and got 17 answers. 20 hours of looking for information were saved over a month. For full-scale deployment, it would involve 300 employees with a targeted time saved of 670 man.days over a year.

Teamcenter PLM support

The company is using Teamcenter as a Product Lifecycle Management (PLM) system. Quite often the engineering PLM support team is answering the same questions to the end-users. On a test month, 2 people asked 2 questions and got 4 answers. 2 hours were saved. For full-scale deployment, it would involve 51 employees with a targeted time saved of 100 days and a 50% reduction in terms of emails.

Other user cases they are considering – not yet deployed: LFE sharing, site intervention and integration within the Knowledge Management ecosystem.

Cegelec CEM ran Proof Of Values to identify the value of Ask for the moon across the whole organisation. The trials took place over a three-month period. The results shown below are monthly averages.

With the AVENIR project – an internal quality improvement cross-company project -, Vinci Energies – the parent organisation of Cegelec CEM – has set itself the strategic objective of sharing and capitalising on knowledge. Cegelec CEM therefore has a strategic objective of sustainable development of knowledge by allowing an increase in the skills of its employees.

Preparing a commercial proposal for a technical tender in the nuclear sector

The commercial team and technical experts write together their technical capabilities based on past projects and documents.

For example, in one tender, 4 people were involved, 3 questions were asked and got 6 answers. 4.6 hours of looking for information were saved over a month.

For full-scale deployment, it would involve 108 employees with a targeted time savedof 52 man.days per year – not correlated with the POV values.

Process optimisation

The internal QA/QC team and the engineering office are writing internal processes and procedures. They identify the gaps and clarify the associated documents based on questions they get from the workforce.

On a specific process, 3 people were involved, 6 questions were asked and got 17 answers. 20 hours of looking for information were saved over a month. For full-scale deployment, it would involve 300 employees with a targeted time saved of 670 man.days over a year.

Teamcenter PLM support

The company is using Teamcenter as a Product Lifecycle Management (PLM) system. Quite often the engineering PLM support team is answering the same questions to the end-users. On a test month, 2 people asked 2 questions and got 4 answers. 2 hours were saved. For full-scale deployment, it would involve 51 employees with a targeted time saved of 100 days and a 50% reduction in terms of emails.

Other user cases they are considering – not yet deployed: LFE sharing, site intervention and integration within the Knowledge Management ecosystem.

The technology: Ask for the moon

One enterprise can own millions of videos, presentations, documents and other forms of information from different data sources. But is it easy to find what you are looking for from that mass of data?

And what about structuring and categorizing the data? If your company has 10,000 hours of video, it’s not really realistic for people to carefully go through and categorize that amount of content. The videos are searchable only by the titles and short descriptions, and the actual video content remains out of reach for the search engine.

Ask for the moon is a Knowledge Management tool relying on social network/ AI/ NPL originally developed for the nuclear industry to answer three main problems:

• Knowledge is a volatile strategic asset;
• Knowledge is scattered across people, teams & tools;
• Knowledge Management is disconnected from day-to-day reality.

Once deployed, Ask for the moon brings the following added value:

• Save time by providing quick answers;
• Derisk projects by providing the right answers;
• Secure knowledge continuously;
• Scale across complex organisations.

 The platform works in three specific steps:

• People ask questions and indicate key tags such as trade, activity/ project or tools involved in the query;
• If the answer is not already in Ask for the moon’s database, the IA finds the right peer to answer;
• The answer provided by the peer, once validated by other users is stored in the database on top of answering to the original question.

When we mention Artificial Intelligence here, we talk about Ontological Engine (identify the links between the different keywords), Collaborative Filtering (to identify the right answer and peer) and Natural Language Processing (to read and analyse sentences).

As you’ve seen across this article, the user interface differs quite a lot from one to another. The reason is Ask for the moon decided to be integrated in daily digital tools used by the workforce. They did not want to be yet another software. They focussed on being integrated in a driven manner.

As of now, Ask for the moon can be used in:

• Microsoft Teams;
• Slack;
• Any mailing software – Microsoft Outlook, Mail, Gmail, etc.;
• Google Chrome.

Moreover, it can be used from a laptop, PC, mobile phone, tablet. The questions can be asked as texts but also through videos, making it convenient when on site.

In terms of deployment, similar to most digital solutions, there is a deployment period to estimate the ROI and ensure the system is suitably working for everyone. On average, it takes less than a month, no matter the size of the organisation.

The deployment phase has three major milestones:

• Week 1: They adapt their system to have a be-spoke AI able to link the different needs and experts from one another within your organisation. The complexity of the company does not impact the duration of development;
• Week 3: Designated catalysts within the organisation are selected and are linked in pairs to engage the workforce in the use of Ask for the moon;
• Week 8: Estimation of the return on investment – ROI. This ROI is estimated both for those asking the questions – time saved looking for an answer – but also those experts giving an answer – not constantly repeating themselves.

On average, deploying Ask for the moon brings those quantifiable benefits:

• 2+ hours of looking for an answer saved a week;
• An increase in reliability as good practices and correct answers are better shared across an organisation;
• -20% of emails sent by those experts who know the answers. They do not have to constantly repeat themselves;
• Positive impact on onboarding of newcomers and better capitalisation of experts’ knowledge.

After deploying the solutions across clients such as Framatome, Cegelec CEM, Airbus, Beaudrey, Daher, Spie , Ask for the moon derived 4 key areas where their Knowledge Management platform brings strong value:

• Operational Excellence: Capitalise on good practices, methods and knowledge. Reduce incidents and anomalies;
• Research and Design: Each project has a particular technical context and the right experts are difficult to identify;
• Commercial Excellence: Meet business KPIs by accelerating the exchange between the teams;
• Complex Projects: Spread the correct answers and good practices between teams in the same project or across projects.

Any questions ? Interested in another innovative enhanced AI-based implicit knowledge platform? Do not hesitate to contact us directly, we will help you find a fit-for-purpose, cost-efficient enhanced AI-based implicit knowledge platform to answer your challenge.