How AI can improve product performance, cut costs and decrease time to market

9 February 2018

My name is Oscar Sverud and since a few weeks I am a proud member of the Imagimob team. As a new member of the team I would like to introduce myself and my experience from managing complex product development projects. Over the years I have been working with several project management philosophies and will below share some of my conclusions in this area. AI is as one of the key enablers to create products that cuts effort in product customizations and boosts product performance, but more about that later. First, I will give some background and problems product developing companies are facing when running complex development projects.

In many companies working with waterfall-based development, big plans are made in the beginning of a project. You spend a lot of time and effort trying to theoretically detail all development that is going to take place the upcoming year/s in detailed specifications. During the development engineers are interpreting specifications with no or very little contact with sales and marketing. Whenever you finish your project, often way past any deadlines set up, the product you have created may in worst case not be relevant any more, or at least is often not what sales / customers had hoped for.

In such projects, sales people and other stake holders are kept in the dark during development or at least this is their feeling. It is not transparent what is happening in the engineering corner. This almost always leads to sales and other stake holders to ask engineers to implement this and that on a daily/weekly basis. Hence, developers get disturbed frequently and priorities change as quickly as the frequency of sales meetings occur. However, this tend to calm down the sales people, they see that their input is getting listened to and they feel comfortable with the situation. The change of focus, disturbances and change of priorities slows down the development to a fraction of the potential. The development is now pure agile in a sense that we change direction several times a week depending on what the latest “most important” customer told sales.

In such an organisation, going agile, e.g. start using scrum or other agile methodology, will to the rest of the organisation (read sales, marketing) seem like a rigid development model. In the past they could always hog the closest engineer and ask them to do this and that super important change. Suddenly when agile is implemented they are kindly asked to go to the product owner to discuss their change. The product owner will prioritize the need and bring the idea up for discussion in a grooming session with the development team. Maybe a few weeks later the change will occur in a sprint. If the benefits of agile principles are not properly marketed to the rest of the organisation, they will experience the change as the development organisation becomes rigid. No more quick fixes and last day customizations. Of course, the power of agile development in speeding up development and increasing quality will vastly outdo the benefits of being able to micro change the course of development daily.

Many final products need customer adaptations. At many companies experienced engineers end up working with tweaks and adaptations of products instead of focusing of the next line of products in development. For example, many tier 1 suppliers to the automotive industry are caught up in endless adaptation projects. Each OEM require that the product is customized to fit their specific brand look and feel. The main selling point for car brand A is that the look and feel is authentic. A surprisingly large part of cars parts is sourced from companies that deliver parts to many car brands. Agile development cannot avoid the need for this type of adaptations. However, when using products that can learn new behaviours, the car OEM will get the power to both easily adopt the product for their specific needs as well as discriminate and create that look and feel that makes them special. 

A current example of this type of AI enhancement of a product is the Imagimob – Autoliv joint project creating a Learning Intelligent Steering Wheel. Autoliv is a tier 1 company delivering safety technology to all major car OEM:s and one of the world’s largest producer of steering wheels. All car OEM:s require Autoliv to customize the steering wheels produced after their specific designs. The learning intelligent steering wheel is of course no exception. The steering wheel is equipped with AI from Imagimob to classify and learn gestures. Since the AI can learn new gestures, the car OEM:s will be able to design their own specific gestures that fits with the car brand look and feel. Engineers will use much less time adjusting for different user scenarios. The cycle for implementing new gestures, trying them out in the field and adapting them till be cut from months to days. Adaptations and time to market with different versions will drop dramatically. Engineering efforts will go down and car OEM's can focus more on what is their core business, that makes the customer buy their car, the brands specific look and feel. 

Would you also like to free your engineering resources from customer adaptations and let them focus on inventing new products? Contact Imagimob to find out more about how AI can boost your product development.