Software-defined products are changing what manufacturers’ engineering departments must deliver today. And as development cycles grow ever shorter, the products themselves become more complex.  

These developments are putting pressure on engineering operating models and tool sets. Engineers now report spending roughly half their workday on documentation, reporting and information searches rather than core engineering tasks. According to Accenture’s estimation, this represents a $39 billion annual loss in productivity globally. 

The following five engineering tenets help manufacturers adapt their engineering capability to the demands of software-defined products: 

1. Stop running the V-model like a relay race. 

Tracking a product requirement from initial idea to specs, to hardware, to code has always been essential to the engineering process. With software and data increasingly defining features and performance, requirement of traceability is more important than ever. Doing the job and then handing over the baton to the next guy no longer cuts it as manufacturers need to constantly provide updates, upgrades and fixes. 

BMW shows what happens when traceability becomes an operating capability. The car manufacturer needed faster access to data from its development vehicles and a better way to track signals across prototyping cycles. It built a Mobile Data Recorder solution that monitors and records more than 10,000 signals twice per second across a fleet of 3,500 development cars. The company then added an AI copilot that lets engineers query this vehicle data in natural language. Now the solution delivers data to engineers and supports analysis about 10 times faster. 

2. Automate verification and compliance. 

A linked challenge is engineering verification. Demand is rising faster than testing capacity, which makes validating and verifying frequent software releases, new hardware and compliance in regulated or safety-critical environments a top challenge. 

Tier 1 automotive supplier Daiichi Elektronik demonstrates the advantages of verification and compliance running inside the development flow rather than at the end. As automotive manufacturer expectations moved toward software-defined vehicles, Daiichi Elektronik had to raise the standard of proof without slowing development. It used PTC’s Codebeamer to establish a workflow linking requirements, tests, defects and releases, which reduced the core implementation effort from nearly a year to two months. With traceability and defect tracking integrated from the start, teams found issues earlier and reduced the rework that often appears with late-stage evidence assembly. 

3. Move to simulation-first development. 

One engineering leader told us, “When we build a circuit, we may have to rebuild that circuit three to four times, and it takes two to two and a half months each time. If we get it right the first time, we save almost half a year’s worth of work.” AI-enabled process planning, simulation and optimization are expected to deliver at least 20% productivity improvement. 

Siemens Energy demonstrates the benefits of simulation and virtual product development. To build hydrogen-capable gas turbines and reduce CO2 emissions, it created an engineering digital thread that integrated all product data and engineering tools like Siemens Teamcenter and Simcenter. This helped engineers visualize, simulate and optimize turbine designs globally across engineering teams. As a result, Siemens Energy’s engineering team completed 26 design iterations—four times more than with traditional methods—and delivered the world’s first 100% hydrogen gas turbine system in record time. 

4. Let AI enable how engineers work. 

Product engineering is quickly becoming an AI-enabled profession. The technology and tools let routine work shrink, and instead, engineers spend more time on judgment, creativity and problem solving.  

ABB Robotics provides a concrete example of that shift. In RobotStudio, its simulation tool used by more than 60,000 engineers, ABB Robotics integrated NVIDIA Omniverse libraries so teams can design, program, test and validate automation cells before deployment. This approach delivered 99% simulation-to-real correlation, cut deployment costs by up to 40%, sped time to market by as much as 50% and reduced setup and commissioning by up to 80%.

5. Unscramble the partner collaboration. 

Product development takes a village. As one senior aerospace engineer manager told us, 70% of an airplane is built by the supply chain. Yet many programs still manage partners through email, inconsistent data and late integration tests. That approach fails when suppliers design major subsystems and when evidence must stay intact across organizational boundaries. 

When collaborating with external partners, engineering departments need to set clear roles, data-sharing rules and access controls from the start. Everyone needs to use the same approved requirements, interfaces and change records, instead of constantly reconciling different files. Controlled digital environments help partners collaborate safely, protect intellectual property and spot integration issues earlier. Standard onboarding, checklists and escalation paths make it easier to bring suppliers into the process without slowing work down. AI can support this by classifying data, flagging oversharing, identifying interface conflicts and routing evidence to the right people. This speeds co-development and validation and leads to fewer late-stage problems, as projects can move only as fast as the slowest partner.