The key to unlocking greater machine productivity and efficiency now lies not in mechanical improvement – but in developments in software, says Alan Berger, OEM veteran and CEO of the on- and off-highway vehicle consultancy, abcg AB
Software on machines is far from being a new phenomenon – but with the arrival of AI and machine learning, it is fast approaching a pivotal moment. Electronic systems are now enabling the mechanical aspects of machines to be optimised and designed in remarkably innovative ways, with some containing up to 10 million lines of code. This level of complexity brings advantages. By shifting more logic into electronics, hydraulic circuitry can be simplified, reducing mechanical complexity and improving reliability. Integrated automation features, such as 2D and 3D control on excavators, enhance precision, lower operating costs and increase operational speed. Machine intelligence also enables better coordination of vehicle subsystems – such as adjusting engine parameters pre-emptively in response to heavy loads rather than reacting afterwards, leading to smoother power delivery and significantly higher productivity.
“An industry-wide diffusion of software knowledge throughout entire organisations is needed”
Furthermore, automated control systems are replacing manual adjustments, ensuring consistent performance across operators and extending efficiency over longer workdays. These advances make average operators highly competent – and allow experienced ones to maintain peak performance throughout their shifts. This shift puts software capabilities as the key source of competitive advantage among OEMs.
From isolated expertise to integrated knowledge
Think back to the last major technological shift in our industry – the introduction of hydraulics in the 1950s. More than half a century later, the industry has a deep understanding of hydraulics – engineers know how to design effective systems, service personnel how to diagnose faults and purchasing departments how to evaluate supplier quality. We need to achieve this same level of vertical integration across the supply chain with electronics and software.
Currently, and understandably, this depth of knowledge is not ingrained in most companies. Some OEMs have capable software teams and the latest tools and processes – but these teams often function separate to vehicle engineering teams rather than being integrated within them. This separation means vehicle teams miss opportunities to improve performance using software or, conversely, sometimes delegate tasks to software teams that should be handled differently. That said, establishing a core of knowledge and solid processes is essential, and I built exactly these stand-alone and holistic software organisations in the engineering teams I led. While necessary then – it is no longer sufficient.
Now, an industry-wide diffusion of software knowledge throughout entire organisations is needed. This means vehicle engineers must understand what software can achieve and – equally important – software engineers grasp the mechanical parameters the vehicle operates under. Service personnel must also properly understand what the software is doing, and why. Purchasing departments should recognise what makes a good software partner, as many components will include both hardware and software elements. Salespeople must become as proficient in software-controlled features as they are in the machine’s physical aspects. And finally, leaders must learn when the infamous phrase that makes software teams cringe: ‘We can fix it in software’, is appropriate – and when not.
A software-defined future
This more joined-up state of affairs can lead to innovative machine architectures that can be optimised beyond current products, bringing even more value to customers. This will essentially deliver the perceived panacea of ‘software defined vehicles’. However, this evolution requires every organisation to go through a step-by-step process, and skipping steps will result in slow progress and frustration.
This article first appeared in the March issue of iVT
