Hello Conrad
For context, and for our other readers, let me just mention that Cisco acquired Viptela Inc., an SD-WAN company based in San Jose in 2017. Cisco’s SD-WAN solutions are based on Viptela’s technology, and terminology, which includes the concept of a Transport Location or TLOC.
Taking a step back, we can say that Cisco’s SD-WAN solution is based upon the use of the Overlay Management Protocol (OMP) which is a routing protocol similar to BGP, which is used to manage the SD-WAN overlay network, which is also referred to as the SD-WAN fabric. This fabric is the software defined overlay that runs on top of standard network transport services and infrastructure such as MPLS networks, or the public Internet.
Getting closer to your question 
a TLOC is used to define a specific interface in the overlay network of an SDN implementation. It is actually used to tie an OMP route to a physical location. It is the only entity of the OMP routing domain, that operates on the overlay, that is “visible” to the underlay network. As such it must be reachable via routing in the underlying network. A TLOC contains various attributes, including IP address, color (an attribute that identifies the TLOC type) and Domain ID to name a few. In other words, a TLOC is a transport attachment point for the overlay tunnel which is used by OMP for routing.
TLOC extensions are used when you want to achieve redundancy across multiple topologies allowing a dual homed design. Similar in functionality to protocols such as VRRP, TLOC allows for redundant connectivity without the need for the redundant devices to be directly connected, such as is the case with VRRP and HRRP for example. An excellent example of how this works can be found on pages 9 and 10 of the following Cisco white paper.
For more detailed information about TLOCs, and SD-WAN in general, the following Cisco documentation is also very helpful.
I hope this has been helpful!
Laz