8 ms·
Launch HN: Terusama (YC W20) – We help warehouses schedule trucks
Hi HN, Andrew and Chris here. We’re building Terusama
(https://www.terusama.com https://www.terusama.com), a truck appointment system for logistics
teams at warehouses and distribution centers. We automate the
labor-intensive process of coordinating arrival times with a trucking
company, checking a truck in at a facility, and keeping track of
everything. Oh, and we have a theme song:
https://www.youtube.com/watch?v=fyp4O4s5tk4 https://www.youtube.com/watch?v=fyp4O4s5tk4
The status quo here is really antiquated software that is really hard
to use. Not only is the current way of doing things inefficient for
logistics folks, it has a big impact. The inefficiency created by poor
supply-chain coordination costs $30B a year in the U.S. and generates
billions of tons of C02, mostly from trucks idling in parking lots.
On top of that, it has a huge impact on the lives of truck drivers,
who are legally limited to 11 hours of driving time each day. Waiting
at warehouses often eats up 3-4 of those. Those wasted hours really
add up, making the job unsustainable for many and contributing to
another problem in the industry - driver turnover, which is over 90%.
Most of this inefficiency is caused by the high level of friction in
communication between the people who manage trucks, the drivers, and
the people who the drivers are delivering to. The freight industry has
made a ton of progress with software to help manage their fleets, but
warehouse logistics software hasn’t caught up.
Chris and I were students together at Indiana University. We became
really interested in this space at our later jobs. I was working
at Uber Freight (formerly Otto), and Chris was consulting for companies in
the industry. We were constantly amazed at how big and complex an
issue it was, and surprised that nobody was thinking about it in the
same way we were. We started talking to as many people in the industry
as possbile to figure a better way to manage logistics. The result was
the version of Terusama we launched with in the beginning of the year.
We have been since iterating on the product with a core group of
customers.
We streamline or automate most of the repetitive and administrative
processes needed to keep trucks flowing smoothly in and out of a
facility. We provide dispatchers at trucking companies and brokerages
a way to search through facilities, and schedule directly with
these facilities. When a truck driver arrives at a warehouse, we check them
in at a kiosk. Our customers get a dashboard where they can track and
manage loads coming to their facility. Logistics teams get visibility
into their incoming trucks, the ability to communicate with drivers
and carriers, and to track a load through its lifecycle (from staging
through completion). Everyone involved is kept in the loop with smart
notifications. In the future, an API will make it easy for anyone with
a brokerage or trucking company to get facility availability (to feed
into truck routing optimization), and schedule appointments. We’ll
also offer services to help carriers reduce the risk of missing truck
appointments.
One of the most important things we’ve learned in this time is how
important reliability is for a mission-critical app. It's obvious in
retrospect, but it definitely changes your approach when building an
application - requiring heavier testing, redundancy, and support as a
form of redundancy.
These days you hear a lot about things like autonomy and AI in the
freight industry. We believe that to get the value promised by these
things, we first have to upgrade the outdated technology
infrastructure that currently runs the industry. Our goal is to build
that infrastructure, while solving today’s pressing problems.
We’re excited to see what HN thinks of our platform. We’re eager to
hear about HN users’ experiences, ideas, and know there is a ton of
expertise among the community to learn from. And if you’re in the
industry, we’d love to hear about your experiences with truck
scheduling!
- hef19898 7y agoAbsolutely agree on the AI vs. legacy systems part! Something most people tend to underestimate when talking AI and logistics / Supply Chain Management. back the day, I helped to roll out a similar, rather primitive system, at Amazon across their german fulfillment centers. Just the added visibility of that was mind blowing. Also increased pressure on operations, I don't consider this a bad thing. If you want to chat, more then welcome! also I'd love to hear about how you approach that problem. Because it is huge one, also one most companies totally ignore.
- adenta 7y agoWe would love to chat! I'll reach out to the email in your profile.
- parthi 7y agoCongrats on the launch! Love that you're tackling an underserved industry
- parthi 7y agoAlso theme song is on point
- dhruvkar 7y agoHi guys, kudos on the launch! I help run a distribution company and I'm not clear who this is for? Is it for my own trucks i.e. scheduling trucks between our locations and/or customers? Is it for when we use a 3rd party logistics company i.e. scheduling time with them to come pick up and deliver our goods? Or is the product FOR the 3rd party logistics company that is handling our goods? Is this usable without two (or more) of the parties using this software? Would love to hear more. There are many issues in this space and love the fact newer startups are getting into it.
- adenta 7y ago> I help run a distribution company and I'm not clear who this is for? Hello! We sell to you, the person who runs a warehouse/distribution center. We are trying to augment the functionalities of a receiving coordinator. Instead of using the phone or email to schedule appointments at your distribution center, you can use Terusama. What can we say to make this more clear in the future? > Is this usable without two (or more) of the parties using this software? Could you clarify your question? We are not a TMS/WMS. We have one customer who still scheduled appointments manually, but uses our kiosks to check truck drivers in & out. Glad you are as excited as we are!