Hello fellow automotive aficionados, engineering enthusiasts, and tech-savvy readers! As a passionate advocate for blending classic automotive charm with cutting-edge technology, I embarked on an ambitious project this year: converting my own 1958 Mercedes 220S into a state-of-the-art electric vehicle (EV). Join me as I dive into this exciting endeavor, armed with a toolbox, a healthy dose of humor, and a vision to create the electric classic car of my dreams. The project’s timeline is set for completion within 2024, working nights and weekends. And perhaps the most groundbreaking aspect? My chief engineer is none other than ChatGPT 4.0. I’m leveraging this powerful AI for everything from in-depth research and complex coding to intricate calculations – tasks that would typically consume months are being accomplished in mere hours.
Consider this your invitation to ride shotgun on this electrifying journey. Think of my AI companion as generative AI with an engineering pedigree. To bring this vision to life, I’m partnering with Fictiv for sourcing and crafting custom components, and enlisting the support of some dedicated friends on weekend wrenching sessions. We’re calling this project “Electra”.
For over a decade, I’ve embraced annual build projects to not only keep my mechanical engineering skills sharp but also to passionately engage with hands-on creation.
Alt text: Dave Evans, CEO of Fictiv, proudly displays the exhaust pipe removed from the 1958 Mercedes during the initial stages of its electric conversion.
Past Fictiv projects include building a miniature electric car, developing an open-source motorcycle, and even undertaking a Fiat EV conversion. Each year, I challenge myself by immersing in a new discipline, dedicating the year to mastering its intricacies.
This year, my passion for climate technology and the capabilities of ChatGPT converged, inspiring me to convert a classic car to electric power. Marrying my early career in automotive engineering with my current fascination with AI, I posed a compelling question: “What if I harnessed ChatGPT for the heavy lifting in engineering? Could we take a vintage car, perhaps something iconic from the 1958 mercedes era, and transform it into an electric vehicle? With ChatGPT as my lead engineer and myself as the mechanic, could we successfully convert a classic 1958 Mercedes to run purely on electricity, eliminating fossil fuels altogether?”
Selecting the Perfect Classic: Why a 1958 Mercedes?
Choosing the right car for an EV conversion demands careful consideration. While many EV conversions utilize classic Fiats, VW buses, or VW Beetles, and sports car enthusiasts often opt for Porsches, I was looking for a unique classic car conversion, something beyond the typical VW projects. I envisioned a sophisticated town car, a vintage chauffeur-driven vehicle – a classic embodying style and elegance over sheer speed, perfect for a date night with my wife. My quest began with a simple query to ChatGPT on my phone.
Alt text: ChatGPT interface displaying the initial search query by Dave Evans seeking recommendations for a classic car suitable for electric conversion, considering factors like headroom and trunk space.
I engaged in extensive dialogues with ChatGPT, refining my criteria: “Suggest a classic car with ample headroom – I’m 6’5″, so legroom is also crucial. I need substantial battery storage, but a usable trunk is still a must.” This iterative process led me to the Ponton series, a line of distinguished 1950s classic cars. The aesthetic appeal, size, and graceful silhouette of the late 1950s Mercedes 220S resonated deeply. The 1958 Mercedes 220S, in particular, emerged as the ideal candidate for this ambitious electric conversion project.
Alt text: A pristine 1958 Mercedes-Benz 220S, showcasing its elegant design and classic lines prior to undergoing the electric vehicle conversion process.
Engineering the Electric Heart: ChatGPT’s Role in Conversion
This is where the project took a serious turn towards engineering. The challenge was to understand the original engine and define the specifications for the EV conversion. I tasked ChatGPT with providing detailed specifications of the existing internal combustion engine (ICE), including torque curves across various RPMs, and gearbox ratios – essentially, all the standard performance data of the 1958 Mercedes 220S. Next, I explored options for electric vehicle equivalents, specifically motors and component sourcing. ChatGPT delivered specifications on a range of motors, from Tesla Model 3 units to Nissan Leaf motors, and options available directly from manufacturers.
Alt text: Graph comparing torque curves and required motor output for different gears, RPMs, and vehicle options, generated by ChatGPT to aid in selecting the appropriate electric motor for the 1958 Mercedes conversion.
Alt text: Chart visualizing vehicle speed across different gears, produced using Python code generated by ChatGPT, assisting in the engineering analysis for the 1958 Mercedes electric conversion.
A crucial decision point emerged: should I retain the original transmission and gearbox of the 1958 Mercedes, or opt for a new system? This choice significantly impacted the project scope. Consider the RPM range: a Tesla motor operates from zero to 18,000 RPM, while the original 1958 Mercedes ICE engine redlines around 5,000-6,000 RPM. Integrating a Tesla motor would necessitate at least a 10:1 gearbox ratio to step down the motor’s high RPM to suitable wheel speeds. This would mean a new gearbox and drivetrain, exceeding the timeframe of a one-year project. The solution was to find an electric motor with an RPM range similar to the original 1958 Mercedes engine, with comparable torque curves, ensuring compatibility with the existing gearbox. The goal was to navigate San Francisco’s hills and enhance performance relative to the original ICE engine, all without overhauling the drivetrain.
ChatGPT then generated Python scripts comparing torque curves across gears for the original ICE engine and various electric motors. These scripts graphed torque curves against engine RPMs to determine wheel speed. While some refinement was needed, the foundational code was remarkably accurate.
In a single Sunday afternoon, ChatGPT produced Python scripts that would have taken me at least a month to write in MATLAB during my engineering student days – and considerably longer now as a weekend engineer. Vehicle dynamics and engine performance calculations aren’t my daily focus anymore. Developing the boilerplate, writing the Python code, integrating necessary libraries, and generating visualizations to achieve the results ChatGPT delivered would have been a significant time investment. In just an afternoon, we evaluated five or six different electric motors, ultimately narrowing down our choice to the HyPer9 high voltage motor – a truly remarkable feat!
Navigating the Road Ahead: Engineering Challenges
Currently, I’m in the phase of tackling significant engineering challenges alongside my friend William Burke, CEO of Five Flute and fellow “super nerd.”
Here are the four key engineering hurdles we are addressing:
- Adapting the new electric drivetrain to the 1958 Mercedes chassis and transmission system.
- Ensuring reliable torque delivery from the electric motor through the stock transmission.
- Designing a robust, waterproof battery enclosure while prioritizing safety.
- Mounting the new electric powertrain while preserving the original vehicle dynamics of the 1958 Mercedes.
We plan to share more in-depth content, host live webinars, and engage with the community to gather feedback and insights as we progress with Project Electra.
Stay tuned for further updates on the Electra project, the engineering challenges we encounter, and the solutions we develop as we breathe new electric life into this classic 1958 Mercedes.
