Increased Efficiency
and Reduced Costs
85%
Reduction in
infrastructure costs
>$140k
Engineering cost savings
each year
150 hours
Saved annually in data
review per engineer
Challenges
Parallel Systems is building a modular vehicle system that enables short-haul and point-to-point rail movements. This specially designed vehicle has the ability to move standard dry van and refrigerated freight through the United States’ existing rail network, carrying a shipping container that can be unloaded with today’s standard lifting equipment.
Parallel started with an in-house telemetry stack comprised of open-source tools. As the company began to scale, it encountered multiple challenges with visibility, storage, data review, and reporting. Built on RDS Aurora, the system worked well when the team and scale of operations was small—with just ten engineers and one vehicle. Though it seemed like a good option built on open-sourced usage-based tools, the system was outgrown when usage was scaled.
Wasted time
Engineers scaling databases rather than building machines
Storage Restrictions
Storage capacity that would become untenable as the company grew
Visualization
Slow and dated UI that impeded insights
Reporting
Laborious manual reporting process
Building a more robust telemetry system wasn’t a core business goal, and the costs associated with maintaining this system were beginning to grow. Parallel had assigned multiple engineers to manage the data pipeline and keep it running, but it was also dealing with the need for more storage and compute.
Migrating data took up too much time, and fixing this system would have required a massive effort, investing additional engineer hours and cutting back on the amount of data they were sending. The team even hired an outsourced developer to chip away at feature requests while the core team focused on building machines.
Reducing data wasn’t an option, and devoting more engineers to deal with the system didn’t make sense. Parallel needed tools that would give it enhanced visibility into the control path of its vehicle, and building these tools in-house required dramatic changes.
Parallel also needed to improve efficiency in the reporting process. In order to bring its product to market, Parallel must work closely with regulators to establish that the novel concept of autonomy meets or exceeds the standard of safety required of conventional freight rail. This requires producing a detailed body of evidence consisting of analysis, testing and demonstration evidence. Creating these reports using common business tools was cumbersome and time consuming.
Partnering with Sift
In order to solve the aforementioned obstacles and streamline development, Parallel partnered with Sift to gain the advanced mission critical tools it needed to progress. Sift’s observability stack provides cutting-edge features that prepare the company for the next set of challenges ahead — so it can continue building, testing, and scaling its vehicles for market.
Data Review
Parallel’s engineers are constantly revisiting its data in order to improve vehicle design. They need advanced visibility into the hardware to monitor its behavior, and the ability to analyze data for reliability and diagnostic purposes.
Leveraging Sift’s fully built-out data review workflow, Parallel is rapidly diagnosing areas to be improved, while giving engineers long-term views on all aspects of its vehicles.
.jpeg)
.svg)
Storage & Retention
Parallel is generating an enormous amount of data which can be expensive to store. It also needs to be able to keep all of that data for testing.
Parallel reduced its database by 85% utilizing Sift’s storage infrastructure – while also speeding up decision-making with faster search times. Engineers can generate all the data they want, only retain what’s important, and continue to generate longitudinal insights over time.
.svg)
Visibility
Developing a new kind of vehicle requires detailed tests that can observe how it’s behaving and identify areas that need improvement. As Parallel continues to scale, it needs the ability to analyze the behavior of an entire fleet of vehicles over time.
Engineers are employing Sift’s Rules feature to confidently discover issues and avoid the risk of missing anomalies. Sift enables Parallel to understand their vehicle performance over time, with detailed and long-term analysis of each vehicle in just a few minutes.
Reporting
Parallel is working closely with government regulators and customers, and it needs the ability to customize that data for each audience. Regulators want a clear and accurate picture of the vehicle’s performance to ensure safety, and customers want data for the vehicles they own.
Instead of devoting multiple engineer hours to manually creating these reports, Parallel is using Sift to automate this entire process from data review through report generation.
"With so much riding on these machines, every engineer diverted to fixing our dated telemetry system feels like lost momentum. We needed a modern solution to help move us forward.”
Kevin Limburg
Manager of Validation and Fleet Software
.svg)
Additional Features
Partnering with Sift gives Parallel a constantly improving observability platform – without the burden of building this kind of system in-house. Parallel continues to bet on Sift as a strategic partner, thanks to continually evolving products that save engineering time and focus.
As Parallel scales its operations, it will utilize Sift functions like data review alerts, run-over-run comparisons, and more customized reporting to further optimize its workflow. Additional tools can help Parallel capture data and statistics for fleet operations and predictive maintenance.
.jpeg)
"What would it take to build and maintain a comparable system? Or even a minimally viable product? It all comes back to core competencies and the value that we get.”
Kevin Limburg
Manager of Validation and Fleet
Software
Faster development cycle
Parallel Systems’ engineers can quickly evaluate simulation and hardware tests, easily share results with their team members, and iterate on simulations to validate fixes. They can identify and overcome a larger amount of obstacles, thereby decreasing issue resolution timeline, and speeding up the development cycle.
Faster development cycle
Parallel Systems’ engineers can quickly evaluate simulation and hardware tests, easily share results with their team members, and iterate on simulations to validate fixes. They can identify and overcome a larger amount of obstacles, thereby decreasing issue resolution timeline, and speeding up the development cycle.
Improved Data Insights and Decision Making
Sift provides deeper visibility into vehicle performance, enabling faster and more informed decision making. Parallel can now generate custom reports for customers and regulators with ease.
"We’re excited to have a front row seat as Parallel Systems enhances the freight industry and solves some of the world’s biggest supply chain issues. Sift is proud to support Parallel’s growth as it brings this game-changing product to market.”
Karthik Gollapudi
About
Parallel Systems (Parallel) enables railroads to serve new markets by converting freight from trucking to rail. Through the use of autonomous, battery-electric, freight rail vehicles, Parallel Systems is seeking to expand railroad service with a flexible, zero-emissions option for moving freight.
.jpeg)
