Excavation on Earth sometimes uses explosives to break up rock or densely-compacted soil. This can be less expensive than using a drill bit or other grinding or impact tools if the explosive is cheaper than the cost of wear on the drill.
On Mars, drilling and grinding tools shipped from Earth are enormously expensive. They could be made from local materials, but not easily and not as an automated process without significant advances. Explosives are in the same boat; anything shipped from Earth is super expensive. Nitrogen is about 1% of the Martian atmosphere, but in a form that requires substantial chemical processing. (Nearly all industrial explosives use chemicals with nitrogen bonds as the source of their explosive power).
One resource that is plentiful is carbon dioxide, CO2. This can be collected directly from the atmosphere and frozen into dry ice pellets. The expansion ratio of CO2 is 845 (1 cc of dry ice forms 845 cc of CO2 gas at standard conditions). A small drill or punch can be used to bore a hole into the work face; CO2 pellets are pushed into the hole and a metal rod with a heater on the end plugs the hole. Heat is applied, sublimating the dry ice into gas and building up a lot of pressure quickly. Used properly, this will cause the surrounding soil to lose cohesion and fragment into clods.
I don't think this would work well on solid rock but it could be very effective on hardpan or other tightly-packed soils where a grinder or impact tool would endure a lot of wear. This could be useful in particular for excavating a habitat shelter; if the equipment can excavate to bedrock using CO2 fracturing and simple bucket or augur tools then the required mass, power and spares can be minimized. Another place where it could be useful is if the surface soil layers hold less water than expected for ISRU operations. Surface soils are loose and easily scooped up with a blade or bucket, so the harvesting equipment will be minimal. Adding a CO2 system like this as a contingency would allow the harvesters to dig deeper and break up ice-bearing soils without requiring any heavy equipment. The force of CO2 expansion should be strong enough to break up solid ice as well, so if a solid layer of ice is encountered it would be valuable rather than difficult.
An expanded use for this might be as a drilling rig using gaseous CO2 as the working fluid, similar to the way shallow wells on Earth can be drilled using water and simple tools. If any soils too dense to be carried away by the gas are encountered they can be fractured by gas expansion; the equipment could be designed to do this without needing to back out the drilling pipe (one-way valve at the end strong enough to survive the gas expansion. If actual solid rock is encountered then a small amount of water can be pumped into the well and frozen at depth, using the expansion of ice to fracture the material. This would be a much slower process but considerably more powerful. This type of drilling is different from taking core samples; it would be intended to reach water ice to be melted and pumped out just like a well on Earth.
Presumably the same mechanism could be used in outer planet probes using methane ice (melts at 90.7 K, boils at 111.7 K) or nitrogen ice (melts at 63.2 K, boils at 77.4 K) depending on local average temperatures. These two don't sublimate but would have to be heated through a liquid phase; still, the temperature rise required is only 21 K for methane and 14 K for nitrogen. Probes to moons beyond Jupiter and to Kuiper belt objects might use this as a lightweight method of digging shallow craters and collecting subsurface samples for analysis, using heat from an RTG directly as the power source.
Another way to accomplish the same thing might be simply to use a laser to deliver a pulse of heat to the end of the hole, causing any volatiles to vaporize and expand vigorously. This would require a material sturdy enough to contain the gases while being transparent to the laser, and would require soil with enough trapped volatiles to produce a useful force. It would also need a fair amount of power for the laser.